Taranveer’s Journal

[ScottMcNab]

I am “experienced” only in changing systems too often in the first few years. I think it is only natural initially as the learning curve is so steep that so many new ideas come along very rapidly and as test them on the original “robust” system it is not unusual that these new ideas improve the back test significantly. Nick posted a while back that one way around this is to wait until the end of the year and implement change(s) at that time if they still seems a good idea. I now do that. Sometimes the idea is enough to form a new system entirely rather than improve an old one which is nice when it happens.

[TerryDunne]

FWIW, what I do when I find a new idea is test it on my current systems.

If the improvement is small, I tend to not mess with things. If there is a really significant improvement, I’m more inclined to create a new system using that idea than tinker with an existing one that has met all of the criteria I had set beforehand. Having said that, I stopped trading two systems that only just met my criteria because of an idea that turned into two new systems that were a lot better.

The exception to this is when I first launch a system – real life can sometimes be different from back testing and I am willing to modify the system in its first few months if I’m uncomfortable with how it operates IRL.

[TerryDunne]

Further to this…I’m still considering changing some of the parameters of my MOC short systems in an attempt to make them less ‘volatile’, even though they went live from the beginning of January. So far, no luck…

[GlenPeake]

Taranveer Singh wrote:

How have you guys “improved” an already tested robust system?

For me, I’ve modified/improved/evolved my Mean Reversion systems since I went live with them in 2019… and I’ve glad I did. The results are much improved.

My thinking is, as Markets evolve then perhaps systems need to evolve from time to time as well, whether that is ‘tweaking’ an existing system or adding an additional system or 2 for diversification.

For the shorter term Mean Reversion systems, I review approx. every 6 months. For the longer term, approx. every 12 months.

I completely agree with Scott’s point around the ‘learning curve and new ideas’. With experience, you evolve as a trading system builder and improve and find new ways of implementing fresh ideas. I’ve recently got some from ideas directly from Scott and am close to ‘polishing off’ another MR system… Cheers Scott

Another example is using REALTEST as an additional way of building/testing/developing new systems.

[TaranveerSingh]

Thanks Scott, that’s a pragmatic approach.

Terry, what you said reminds me of the course itself. If realtime results don’t match backtests, there’s a problem. Thanks for the reminder. Luckily, I have monthly systems. 100% with you on when not to tinker.

Glen, I can see the benefits of changing if the results are significantly better.
When you say, “as markets evolve, so should the systems. I do it every 6 months”
Do you see if your parameters are still in the middle of the flatline?

Unless I’m missing something, this is what Nick says about a broken system “it’s broken only if it’s not performing due to reasons you can’t explain.” So that leaves out the systems that got punished in March 2020.

^Keeping this in mind, how do you guys ensure you’re not fixing what’s not broken in the first place?

[GlenPeake]

Taranveer Singh wrote:

Glen, I can see the benefits of changing if the results are significantly better.
When you say, “as markets evolve, so should the systems. I do it every 6 months”
Do you see if your parameters are still in the middle of the flatline?

RE: Review 6 months etc.

This is something that Nick has suggested in the past….

For the shorter term MRV systems I (now) review each 6 months…. e.g. 6 Month Review

RE: Do you see if your parameters are still in the middle of the flatline?
…just looking for shifts in the performance for a parameter, double checking for the ‘sweet spot’ in performance. I like to look back and see how different variations of my systems performed with alternate settings etc.

After going live with my MRV systems I let this ‘review’ period slip out to approx. 12 months….. had I reviewed earlier I may have adjusted a parameter which would’ve netted a positive return for 6-12months. So now it’s something I like to keep on top of….

E.g. Review

[TaranveerSingh]

Before I can share my results for this month, I wish to share what (to me) is more important.
I had 4 versions of strategies running and I didn’t feel tempted to check anything.
Had a hedge on which gave up its gains in 2 days and I am alright with it. Really proud of myself for this bit. Hoping to continue this detachment but I’m sure it gets more testing in a downtrend.

In the first trench, I had some mirroring stock pairs (UAA-UA, LGF.A-LGF.B, NWSA-NWS) so the only rotation this month is to get rid of these.

Trench 1
R1000 monthly momentum = +11.6%
NDX monthly momentum = +0.7%

Trench 2
R1000 monthly momentum = +2.8%
NDX monthly momentum = -0.8%

[RobertMontgomery]

Taranveer, Thanks for sharing your thoughts on detachment. I find it best to be away from the computer when the market is open and avoid watching it on my phone as much as possible. I have stopped tracking my positions on my phone and I find it best when I just wait and see the results at the end of the day. I try to remind myself down days are part of the process just like up days, just trust the edge created by following the process.

[Nick Radge]

Quote:

(UAA-UA, LGF.A-LGF.B, NWSA-NWS)

You can create a watchlist for these and add them to the Exclude list in the filter.

[JulianCohen]

Rob M wrote:

I find it best to be away from the computer when the market is open and avoid watching it on my phone as much as possible.

I agree Rob. Australia is a great place to trade US markets from that perspective. You set your orders, go to bed, sleep through the action and do your accounting in the morning. Nice and boring, with the occasional expletive when you first look at the results. Then it’s back to boring again.

Love it

[TaranveerSingh]

Quote:

I try to remind myself down days are part of the process just like up days, just trust the edge created by following the process.

Thanks Rob, I always try to tell myself that too! Prevents me from doing ridiculous excel extrapolations

Quote:

You can create a watchlist for these and add them to the Exclude list in the filter.

Thanks Nick, yes I’m doing that. Will keep looking out for these going ahead and update the exlude list.

Quote:

Nice and boring, with the occasional expletive when you first look at the results.

Hah! This should replace #next1000trades

[TaranveerSingh]

(off trading)

So I took a trip to Goa and stayed in a hostel. Almost everyone I talked to, traded/invested and when they described how they do so, I felt insanely grateful to the course. It’s almost scary what people do with their hard earned money.

I also got offered a job in sales for a lease financing company as a friend who works there put in a good word for me. Remote job, pays well, aligned with my values. Round one interview went great. I shared my horrible job experiences and then went on to send them a mail just so that they understood the gravity of why I demanded “sanity” at work. For those who don’t know, India is infamous for just thee worst working cultures ever. You’re shamed for being transactional or for wanting a life. In my last job I’d often have a week’s worth work hours as (unpaid for) overtime in a month.

In the email, beside the work hours and off work mental space, I even mentioned that since I started trading , I’ve found myself growing less tolerant of toxic work cultures. I respectfully disclaimed/asked for things I’d have never imagined… and they agreed!

While the round two is due on Monday, I do feel they really want me on board. And this is the first time I negotiated for off salary stuff. Wouldn’t have been possible without the course. I think finding myself in a terrible workplace might be the biggest “drawdown” I think I have avoided? Thanks for enabling this Nick, Craig, Trish and the whole freaking forum!

[Nick Radge]

Great news Taranveer. Interesting culture. Unfortunately I’ve not been in an institutional environment for 15+ years (thankfully) but I certainly couldn’t think of ever working for anyone else anymore.

[TrentRothall]

Great news mate

[TaranveerSingh]

Trench 1
R1000 momo = +8.99%
NDX momo = +0.11%

Trench 2
R1000 momo = +6.22%
NDX momo = +0.97%

3 rotations in the NDX due. None in R1000. Didn’t (have to) know S&P500 was near ATH until today. Still as fascinated by such stuff as I was on day 0.

[Author]

Posts