The momentum anomaly is well documented factor. Various studies have shown a 12-2 months to be best as the -2 takes care of the factor mean reversion. Look at various papers by AQR, Alphaarchitect, Meb Faber etc. Wes Gray has show (in his book Quantitative Momentum) that rebalancing on the last day of the month previous to the quarter end, presents the best opportunity to take into account fund manager window dressing.
Another recent paper by Hoffestein tested the continued alpha post signal and found it to decay within 4 months. The better approach might be a 6-9 month momentum filter and then hold only to month 14 then rebal.
You should also look at Stocks on the Move by Andreas Clenow. He uses a liner regression slope, R-squared over 90 days to project 12m momentum going forward!
For some strange reason I suddenly decided to return to looking at the 12:1 filter this morning. I wrote a DJI weekly strategy that uses two different ’12:1′ filters to decide on market entry. They are not set to 12: 1 but different ratios obviously as 12:1 is for monthly markets. The exit is probably as important if not more important than the entry but I will keep that as my little secret!
The results were rather pleasing. This is with spread of 3.8 and level stake of 1. The maximum draw down compared to the overall profit is very good in my opinion. I think this sort of filter has great potential from looking at these results and is definitely worth further experimenting with. This strategy is definitely one that I will put on forward test.
Vonasi,
I have experimented with this, using my strategies and it’s looking very good.
Looking into the DOW weekly, not many trades, but still massively profitable.
I am not sure how to apply the filters, to check the daily etc?
I am not sure why it hasn’t been talked about more.
See attached a screen shot.
Drawdown is £4k, worst trade is £382, best trade £6,5k.
Vonasi,
I have been playing with this for a month now, and the results are promising.
How did it do, on the forward test? I am about to set it up.
I have set my own exit strategy, as you did, max position is set at 1, when at 3 or 10, the results are superior, but the drawdown is much higher.
For some reason the gross performance is not showing, but please see the attachment.
For some reason, the start date is 4 Feb 2106, not sure why.
How did it do, on the forward test?
I’m afraid that I can’t say how it did because I lost two accounts to sudden IG culls of accounts and so lost all live forward testing results. Due to the IG limit of only 25 strategies on forward test I have never got around to putting it back on live testing. I managed to back up some of my strategies at the last cull so I’ll have to check to see if it was one that I managed to save and then run a back test to see how it fair even if only with a fixed spread.
Vonasi,
Thanks for your reply, the data would be great.
I presume IG don’t charge for PRT on a demo account, I have not used their demo account before, I trade at least 4 time’s a month live.
The weekly is great, I might try to convert to a daily format.
I trade at least 4 time’s a month live.
If you do that then IG will keep you live. I had several accounts so that I could forward test 75 strategies but I only traded properly on one account. I did minimal quick in and quick out trades on the other accounts to meet the four trade criteria but IG still deemed it to be of ‘not sufficient quantity’ and just culled the accounts without any warning or even polite request to increase my ‘sufficient quantity’. Be warned – IG make the rules and interpret them however they decide to!
im getting this when imported?
https://imgur.com/a/ava9GVU
I tried it on a stock, it seems i need to look at an indicy?
Seems to work better now that im looking at US500
It does work much better on the DOW.
Hi Guys,
This Filter is simply a derived momentum indicator based on Open and not Close (That’s why it’s work well) as you can see on this picture 😉
Have a nice sunday
Hello, Great job. What I like is its simplicity and the fact that you can play it on several indices. Is it possible to attach your strategy to backtest it? Thank you for your sharing …
If it is my strategy that you want then please read earlier posts where I explained that it was lost in an IG account cull of under used accounts.
Even if I could find it again then I wouldn’t post it here as I have a personal policy of not sharing strategy codes. This is for two reasons:
- If it turns out that the strategy is rubbish and loses money I would not want to feel responsible for other forum members losing money.
- If it turns out to be a really good strategy then I would not want to share all my hard work and time and effort testing for free.
I understand that this policy of not sharing strategies may upset some (especially those who want to cut corners and put as little effort in to get rich as quickly as possible!) but I like to think that I have made up for this lack of strategy sharing by creating and sharing for free many of my own indicator codes and analysis tools.
Other forum members like to put their codes out on the forum and share development and testing with the community – I just prefer to keep it all under my own control.
