Without any tweaking or adding new Systems to my 94 (43 on my wife’s account) Systems under Demo Forward Test … overall performance of my portfolio has improved significantly in the the last 7 – 10 days (from 26 Nov 18 ish).
Either I have finally got it right with the balance of the various Systems or (more likely?) the Markets are just acting more predictably or something?
Anyone else experienced similar performance improvements in the last week or so?
First of all congratulations GraHal, I guess that the hard work eventually is rewarded (applause!).
I haven´t notice myself these increase in performance these days, are most of your systems of the same type: trend following, mean reverting, breaks….? that could explain it.
are most of your systems of the same type: trend following, mean reverting, breaks….? that could explain it.
I don’t have emphasis on any one type … I’ll go with any that shows robust performance in Back Test and Forward Test.
I have for many months been weeding out under performers and adding better performers so maybe the hard work is paying off, as you say!? 🙂 More than likely a fluke and it will be all downhill again next week!! 🙂
Lately I´ve been thinking that my systems tend to work in a particular way that maybe it´s related to what you are experiencing. It could be called the “leap effect”
Most of the time my systems don´t gain much or loose, you know: loose some, gain some and so on… quite a flat profit curve. But occasionally they gain quite a lot in a few days and these occasional leaps are what make most of the profit in the long run.
Or maybe I just imaging patterns also in the profit curve 🙂
As an easy way to see at a glance how I am doing across 4 Platforms I make the Funds available in IG up to a round figure every so often (after a major sort out of Systems etc) … 100,000 on 2 Platforms and 1000,000 on the other two.
The funds available has always always on average slid downwards, sometimes alarmingly quickly … like 2k per week ish! Yes a few leaps up also but the net overall result has always been that the funds available decrease over time! 🙁
But there has been a turn around in the last 10 days or so … like 1 Platform is showing near £12k profit! As you say maybe it is just a BIG LEAP up and it will be followed by a VERY BIG FALL!!! hahahhah
Just yesterday I was listening to a YouTube trading guru whilst washing up and he basically said that he tried to create forex strategies that were basically just good at breaking even – but he would open two positions and if his target was hit he would close one position and let the other run with a break even stop loss. Most of the time he would break even but every now and then he would hit a home run on a big trend. Maybe this is the ‘leap effect’?
Without partial closure on live PRT it seems pointless testing the theory at the moment but my first thoughts were that for every small win he was doubling his risk with two positions that could lose – so his strategy would have to be pretty good in the first place just to break even with most of the time either two losing bets or one winning bet and one break even bet before he hits the home run. Sometimes these YouTube gurus do talk a good trade but miss some of the little details out like the fact that he needs a really high win ratio to just break even!
Grahal, and this “leap” in your profit it´s wide spread between systems or concentrated in one or few systems?
Vonasi, IMO that´s a quite used techniqued in manual trading and I always thought that it can be good not because of the overall performance but because it manages better the psycological aspect… kind of a dinamic stop taking partial beneficts. So better %WinRate but I not sure if better P/L. I am also curious about how it would result in backtesting with partial closure.
that´s a quite used techniqued in manual trading and I always thought that it can be good
Yes it is a commonly touted idea but my thoughts are that hitting the odd home run would be great – but first of all we have to have a strategy that wins twice as often as it loses and just breaks even or makes a small profit after all overnight fees (the guru was a daily chart trader). Maybe searching for the holy grail constant profit strategy is where we are all going wrong and we just need a high hit rate and poor profits!
this “leap” in your profit it´s wide spread between systems or concentrated in one or few systems?
Well PRT doesn’t make it easy to analyse performance in my opinion? How does anybody else do their analysis?
I sorted my Orders List by System but nothing really jumped out at me!
I sorted by Gain and some results are attached. Contrary to the consensus view, it seems that MA Systems feature heavily in the leap in profit over the last week or so??
Glad you asked as I can see now that my hunch appears correct in so far as most of the gains are by Systems showing a version of the last few months (after my recent weeding out and backtesting / tweaking etc loads of Systems).
Also I noted that the duration of the majority of trades are in hours.
There’s even a System of yours showing TempusFugit … with a gain of $435!
If only the gains were real / Live trading! 🙂
I notice that a most of your big wins are on the DJI in the days just following a big gap up at the week open.
[attachment file=86382]
Perhaps this had some effect on your strategies?
Perhaps this had some effect on your strategies?
Yeah maybe … strategy for a fall following a big rise … often happens and has worked this time?
Don’t be fooled by the recent market behavior, the recent volatility (and very directional intraday!) necessarily have a significant impact on performance. Same thing appears once in a while in backtests if you look carefully 🙂
Same thing appears once in a while in backtests if you look carefully
Yes I agree and I didn’t purposefully set up any strategy for … big rise then fall / big fall then rise … but maybe the variable values picked up rises / falls during backtest / optimisation and set up accordingly.
Hoorah for optimisation!!!! (joke! )
I’m not holding out any big hopes … I’ll keep this thread posted on how it all develops over the coming weeks.
But don’t let that stop any further comments … it’s good to talk! 🙂
I have seen some great performance, but also some horrible losses these days.
Days when things are down 2-3% are normally OK for me, im 8/10 times not in the market then cus everything looks horrible. (im long only)
Days where things are up 2-3% have been amazing and ive seen some trades that actually have outperform any trade in the backtest. Obviously lucky, but i wanna quote a podcast:
“You have to stick it out during the rough times, because thats where the biggest green days are. If you want big gains you have to have big risk.”
And i think that is definitely what has happened to me these past few weeks. The volatility has somehow created bigger opportunities then previously seen in the backtest. So if u just manage to not bleed out and die during the rough parts there are usually big gains waiting.
I hope we’re not going into some sort of 24 months of crazyness. It has been fun running my algos live during this period of volatility, ive had 1 fail miserably but most are actually doing good.
