Still in cash :-)

[GlenPeake]

Thanks for the updates Mike….keep ’em coming… Always nice to read through the journals to see the progress and follow along as the systems take shape….

Are you still looking at a rotational system to start with?

Which market(s) US or AUS?

[MichaelRodwell]

Cool – someone is reading!

Yep, rotational style for the first one. I really needed somewhere to start and focus. Work is pretty busy for me at the moment and requires a lot of travel so the lower admin and operations was appealing.

Luckily my wife is fully on board so when our youngest get a little older she’ll be able to run and maintain trades on a daily basis – worked well in the past when we traded some of Nick’s systems.

Going with the US to start with but a longer term goal is to trade both markets with multiple systems… easy does it for now though!

[MichaelRodwell]

For me, the question of the Index Filter was on death’s door step. I had sliced and diced it every way I knew how and then it came to me.

A new idea.

I’ve spent a lot of time becoming acquainted with the $SPX and the 200MA and I really wanted to figure out a unique way to be in or out of the market.

May be this is it?

More work and validation to be done but I have been able to reduce my MaxDD by approx 3% while reducing my CAR by less than 1%.

Current stats:
CAR: 15.11%
DD: 16.23%
MAR: 93

[MichaelRodwell]

I had a great call with Nick today. Well it was great for me any way!

We spent a fair bit of time looking at my index filter idea and adjusting some parameters for other parts of my system. One thing I learned was that as a general rule choosing parameters of similar lengths is a good starting place to ensure the system is not being curve fitted. For example, using the 200MA as an index filter and using 200 bar length for ATR look back.

Post our discussion I have spent most of the day doing more tinkering and testing.

My promising index filter idea is now on the scrap heap with the others.

200MA reigns supreme for the moment but I have a few more ideas to test.

Would love to hear any feedback on people’s success using different index filters for monthly rotational systems.

[TimothyStrickland]

As you mentioned and I did the same thing as well Mike. My first rotational system I ended up curve fitting. After Nick looked at it we decided I should set all my parameters back to the default and try again.

I am sure Nick told you this already but for me I pick ONE thing and one thing only that changes my system the most and optimize that one parameter only. Figure out what that is, every system will be different, in some it was the index filter, in another it was the Stock Filter moving average, in my mean reversion system it was holding period.

I think it is a good exercise what you are doing, play around with each parameter so you have a good idea what your system is doing. For me I have used index filters as high as 270 MA. I do not think the Index Filter would be your biggest system change but who knows!

[MichaelRodwell]

Thank you Tim!

Great advice there!

I’ve realised there is just no substitute for experience and putting in the time to investigate things even if they probably wont work. Every failed test is a actually a win for me as I discover what doesn’t work and a little bit of trading system design capital is added.

As I said to Nick on the call. I’m under no illusion that I even know what I don’t know.

Regarding the index filter, it’s just where I have started my research. I want to be pretty thorough before jumping on to other things and since I still have plenty of time I’m comfortable with that. If at the end of the process I can’t improve it, I’ll feel good knowing I tried a bunch of things.

That’s the confidence I want to build.

[GlenPeake]

Hi Mike,

If you haven’t already done so, I have a read of this PDF.

https://cmtassociation.org/wp-content/uploads/2015/11/0107-geisdorf.pdf

Trend Following & Mean Reverting Indicators: How to Use, When to Use, and How to Use Together.

This is what led me down the ADX Index filter path…. so there might be something else in there that triggers a light bulb moment for you….. perhaps not directly for your current Rotational system, but for other systems down the track.

Glen

[MichaelRodwell]

Thanks Glenn. I did have a quick flick and took a note to look at trying something with ADX or similar and incorporating it with indicator or method of identifying trend direction. Some of the content is a bit of grind for me still but I have stored that one away in my evernote of ideas to try.

[MichaelRodwell]

I had a good call with Nick yesterday. We looked at my system results and some more areas of investigation to get closer to my goal.

For those playing along at home my goal is:

CAR: 20%
MaxDD: 25%

I’m sticking with my ROC momentum measure for the moment while trying to improve a few other areas of the system.

Today’s massive leap forward was the implementation of a weighted dual measurement of momentum.

This increased my CAR by a percentage point while reducing my MaxDD by about the same.

Pretty cool indeed.

Up to date stats:

CAR: 17.21%
MaxDD: 18.47

Very palatable!

[LEONARDZIR]

Mike,
Which market did you test?

[MichaelRodwell]

Hey Len – this is S&P500.

[SaidBitar]

On the NDX MOMO i used two periods look back for measuring momentum and it works good mainly because the number of positions is small but i did not see big difference on SPX momo where i can hold way more stocks

[MichaelRodwell]

Quote:

Said Bitar: On the NDX MOMO i used two periods look back for measuring momentum and it works good mainly because the number of positions is small but i did not see big difference on SPX momo where i can hold way more stocks

Hey Said. When you say good, what does that mean? I was able to improve my return by about 1% which I was happy with but having some context from someone else would be good. Perhaps my ranking method is flawed…

Have you found that using worst rank held makes a difference for your momentum systems? So far for all my system iterations I havent found a way to use this to my advantage. Thanks!

[TimothyStrickland]

Mike, it may depend on how your system was built. Mine makes a big difference. I think the more positions you are holding the less difference it is going to make (in theory). For instance, my NDX Momo only holds 5 which would be considered an aggressive approach.

[MichaelRodwell]

Pretty happy right now. I’ve managed to get my system returning above 20%. :cheer:

Before I get ahead of myself i’ll need to get the code reviewed and also get Nick to take a look to make sure I’m not doing anything stupid.

I’ve made a few changes:

– Index filter is now using dual moving averages
– Dual momentum weighting shifted
– incorporated the worst rank held function
– changed universe to NDX

alrighty then… load up the account, let’s go! (just joking).

[Author]

Posts