Dump it Here

[Willzy]

Hey all, I have a question that I'm hoping someone on this forum can help me with.

I have a system which trades specific US stocks such as MSFT or VISA etc. The system trades LONG only, holds positions for 1-10 days and is mean reverting in nature. It will trade between 150-250 positions per year. Positions must be taken at the close of the market when the signal for entry or exit is calculated.

Currently I trade the system live using Admiral Markets Metatrader 5 platform trading CFD's on the stocks. Automation is important because of my irregular day job.

Like all mean reversion strategies the system is highly susceptible to commission drag. In my case the size of the spread is roughly 0.1% of the traded volume ie a $10,000 position will cost me $10 round trip plus swap rates for the CFD's. The system uses a small amount of leverage less than a 2:1 ratio so a margin loan account could do the job.

The system is currently profitable with these costs but if I can get the cost of trading down it would be worth the effort in re-writing the robots and changing platforms.

I understand that it is possible to find some brokers who offer $0.00 commission on US Equities this would save me big $$$ But I still need access to automation and if possible some form of leverage...

Possible solutions could be
Interactive Brokers with an API ==> probably beyond my skills as a programmer
TradeStation == > Might be possible but I have no experience with them.
Several Non-ASIC regulated MT5 brokers based in Europe / Cypres ==> could be a bit dodgy

Any suggestions or recommendations or past experence with IB or TradeStation would be greatly appreciated as this particular problem is doing my head in...

Cheers Guys!

 

[peter2]

Accumulate the profit that you're earning from the cfds and pay someone to code an API to use on IB as IB have the cheapest costs in Aust. I'm a TS client and pay a minimum $5ew for up to 1000 shares. The zero commission rates are for US clients only (as far as I know).

 

[Willzy]

Accumulate the profit that you're earning from the cfds and pay someone to code an API to use on IB as IB have the cheapest costs in Aust. I'm a TS client and pay a minimum $5ew for up to 1000 shares. The zero commission rates are for US clients only (as far as I know).

Hi Peter,

Thanks for the heads up with TradeStation They were in serious contention but $5ew is probably no better for the size of account I currently trade. $5 each way will be competitive in the future though!

Are there any other costs with running TradeStation?

Cheers
Willzy

 

[Willzy]

Further to my reply above. Is anyone here running an API with IB and if so did you get someone to build the API or go it alone? All info is appreciated

Cheers

 

[Warr87]

Further to my reply above. Is anyone here running an API with IB and if so did you get someone to build the API or go it alone? All info is appreciated

Cheers

i hired a programmer. he did a great job but i had some issues. not his fault. he was used to doing intra-day so a portfolio level design was different for him. i found him, and a bunch of other candidates, on the freelance website UpWork.

i stoped the papertrading with the API/IB since the TA libraries in python were giving different values to AB. The data was the same from norgate (norgate have a python library). It was a complete mystery to both of us which is why I gave up on it (for the moment).

If you don't necessarily want a python program written up to work with the API, you have options to use the Smart API from The Chartist. Another option is Alera Portfolio Manager. Both of these integrate into IB's API to send orders. The Chartist doesn't offer screenshots or an overview of how it works, and APM offers minimal info and tells you to go to a third party site and buy there course on automating AB with IB. I wish I knew more about these 2 programs. (The Chartist API is also one off payment, APM is an ongoing subscription fee or free with an IB demo account.)

Let me know how you go. I am tempted to go with either Chartist or APM but I also need more info to make a decision.

 

[julo]

My view is less 'rules' and more about risk mitigation. Just need some way to jump on a trend and then ride it. Take out the noise - way easier on weekly charts and Aussie stocks and suffer some drawdowns which are the price of admission. Look for less 'efficient' markets - good luck system trading the S&P500 for example.

Hi @investtrader / Gary,
do I understand your point correctly? You recommend less efficient market because there are imperfections that can be used in our favour? And S&P500 is too effective (and hard to trade) because there are many (clever) sharks in the sea?
Thanks!

 

[investtrader]

Hi @investtrader / Gary,
do I understand your point correctly? You recommend less efficient market because there are imperfections that can be used in our favour? And S&P500 is too effective (and hard to trade) because there are many (clever) sharks in the sea?
Thanks!

Yes, that is what I meant. IMHO if you take a trend following system and apply it to the S&P 500 stocks it generally does poorly compared to ASX stocks.

 

[Warr87]

Yes, that is what I meant. IMHO if you take a trend following system and apply it to the S&P 500 stocks it generally does poorly compared to ASX stocks.

I'm sure that would be true. But a MR system, from my understanding, is better for indexes so I wouldn't be surprised if a trend following system would not perform as well.

 

[Skate]

Skate.

 

[Skate]

Skate.

 

[Skate]

Skate.

 

[barney]

Thanks for your continued input Mr. @Skate

The Cat is happy, The Panda is purring and the Zebra is displaying his stripes for the moment

 

[Skate]

Skate.

 

[Skate]

Find it interesting and maybe an incentive for me to give a go to a back to basic trend system, to add diversification

if you do have that divergence then maybe a basic trend system would be a good idea!

The two quotes by @qldfrog & @Warr87 got me thinking about making a post of a "simple trend trading strategy" - with a twist of course. As trend trading strategies go, using "Guppy's Multiple Moving Averages" would be a suitable choice as it's easy to follow & understand. The relationship between the 12 exponential moving average (EMA) tells the story of long & short term buyers.

The Guppy Multiple Moving Average (GMMA) indicator
This nifty homegrown lagging indicator provides an exciting approach using 12 (EMA). The Guppy Multiple Moving Averages is a great tool if you are considering building a trend-following system. The (GMMA), is a fine technical indicator that identifies the strength & changes in trends. The changes in the trend are indicated by the GMMA ribbon. The combined (EMA) ribbon times the entry & when to get out of a trade.

Multiple Blue & Red Lines (moving averages)
The (GMMA) is composed of multiple lines that help traders see the strength or weakness in a trend better than if only using one or two moving averages or (EMAs). Compression to the expansion of the “lines of the ribbon” tells one story whereas the reverse (expansion to compressions) tells another. On the other hand, the "crossing" of the ribbons, is a whole other story in itself.

GMMA Ribbon colours
The short-term investors are represented by the “Blue Ribbon” & the longer-term investors are represented by the “Red Ribbon”. To keep the post short, those interested can do their own research as to how these lines (of the ribbon) interact. We can use the association of the "GMMA ribbon" (the lines) to our advantage. The downside to (GMMA) - it’s a "lagging indicator" that will never catch the low of the pivot but at times goes very close.

Twisting the GMMA
I’ve twisted Guppy’s idea. I've taken the average of Guppy's slow & fast-moving (EMA's) then added a smoothing factor. Using the average of the EMA bands (IMHO) sharpens the usefulness of the indicator allowing me to turn the GMMA indicator into a complete trend trading strategy.

More to follow...

Skate.

 

[Skate]

Cracking the GMMA indicator
As a trend trader, it’s just not enough to identify a confirmed trend. Timing the entry (when to jump into a trend) is so important. The Guppy's (GMMA) Indicator allows you to get into a trend as soon as possible & to get out after a trend has reversed.

Reading the chart below
I've marked the chart in yellow that holds an interest (the strategy signal area). Other patterns repeated within the chart are considered false patterns. The GMMA ribbon (the indicator) identifies the trend, the strength of the trend & the reversal of the trend. The "blue ribbon" is the short term exponential moving averages (EMA) whereas the "red ribbon" is the long term exponential moving averages (EMA). The "GMMA Strategy" uses the average of the averages. (how's that for double Dutch)

Marked in "yellow"
The multiple lines of the Guppy's (GMMA) help traders identify the strength or weakness within a trend. The "GMMA Strategy" has a buy signal for Monday 14th December 2020 so I'll use that as the example. (EMR)



Backtesting & Monet Carlo runs
I've explained to @Roller_1 recently that I have the basic Norgate data package. Norgate Silver subscription doesn't include "Delisted Equities" or "Historical index constituents" & that's why I normally post recent backtests for this very reason. To me, backtesting gives a base comparison at best, a comparison between strategies or within a strategy. Backtesting & real-live trading 'in my opinion' will never truly align as there are so many variables. Backtesting on historical data alone means "JACK" & should never be relied upon. 'Out-of-sample' data is the only true measure of value, meaning paper trading is the only real measure of a systems worth.

It's worth remembering
You're up against traders who will have more experience, more information & much more money than you, so we need to concentrate on survival when thinking about developing a trading system. Making money in this game is secondary.

Short backtest period
The "GMM Strategy" backtest period is for this calendar year (1st Jan - to - 11th Dec 2020). The results posted below "should never be relied upon" but it gives a clear example that using Guppy's GMMA is worthwhile in coding a simple trend trading strategy. The results are posted in the spirit of sharing.









Summary
The 'Dump it here' thread is about ideas, ideas that I've found helpful in my trading experience. The Guppy GMMA indicator falls within the scope of this thread.

Skate.

 

[Warr87]

For a 'simple' trend following system I was thinking something as simple as a donchian, or an MACD system. But can't argue with your results!

 

[investtrader]

Hi Skate,
Why doesn't the system take a Buy on about 17th Jan? Do you have another rule besides the GMMA signal?
Cheers

 

[Skate]

Hi Skate,
Why doesn't the system take a Buy on about 17th Jan? Do you have another rule besides the GMMA signal?
Cheers

@investtrader the succinct answer to your question is yes & there are other rules besides the crossing of the two ribbons for a buy signal. The GMMA ribbon as shown on the chart was for example purposes so others would know what I was trying to convey. The GMMA ribbons are subjective in interpretation due to compression & expansion of the ribbon EMA lines. The "compression to expansion" & "expansion to compression" of the GMMA EMA ribbon (long & short) tells a powerful story where the price might be heading next.

"Buy on about 17th Jan?"
The reason the position was not taken earlier (on about 17th Jan) was due to the fact the "Buy Condition" wasn't satisfied. The GMMA buy signal of my "GMMA Strategy" is the average of the "6 short term EMA's" crossing over the "6 long term EMA's" with a smoothing factor applied. The buy signal is also at the mercy of filters & parameters. If you look at both charts you can see that there is an "incomplete cross" of the EMA's (the yellow circled area). The 2nd chart below is expanded for clarity







Closeup view of the chart in question
The closeup shows the "incomplete cross" of the EMA's (the yellow circled area). FYI - the "GMMA Strategy" buy condition is "the average of the average" with a smoothing factor applied. I should also point out if you use a low smoothing factor the signals are closer to pivot low but creates a fair amount of whipsawing. A longer smoothing factor reduces whipsawing but at the cost a later signal. It's a trade-off.



Skate.

 

[investtrader]

Okay, I see. You are strictly using a cross as opposed compression of the averages.

 

[qldfrog]

@investtrader the succinct answer to your question is yes & there are other rules besides the crossing of the two ribbons for a buy signal. The GMMA ribbon as shown on the chart was for example purposes so others would know what I was trying to convey. The GMMA ribbons are subjective in interpretation due to compression & expansion of the ribbon EMA lines. The "compression to expansion" & "expansion to compression" of the GMMA EMA ribbon (long & short) tells a powerful story where the price might be heading next.

"Buy on about 17th Jan?"
The reason the position was not taken earlier (on about 17th Jan) was due to the fact the "Buy Condition" wasn't satisfied. The GMMA buy signal of my "GMMA Strategy" is the average of the "6 short term EMA's" crossing over the "6 long term EMA's" with a smoothing factor applied. The buy signal is also at the mercy of filters & parameters. If you look at both charts you can see that there is an "incomplete cross" of the EMA's (the yellow circled area). The 2nd chart below is expanded for clarity

View attachment 116295





Closeup view of the chart in question
The closeup shows the "incomplete cross" of the EMA's (the yellow circled area). FYI - the "GMMA Strategy" buy condition is "the average of the average" with a smoothing factor applied. I should also point out if you use a low smoothing factor the signals are closer to pivot low but creates a fair amount of whipsawing. A longer smoothing factor reduces whipsawing but at the cost a later signal. It's a trade-off.

View attachment 116296

Skate.

Thanks Skate, GUPPY why not, as good as any start base, and better than many.I will play on this concept and try to add a basic system to my portfolio.And thanks for providing a comparison backtest to ensure I am not completely off the chart
Trying to remember simplicity is best