Still in cash :-)

[Anonymous]

Mike, I had some success using something simple like VIX volatility limit in my index filter. It particularly helped in the years around 2009 through 2011. Just something simple like VIX must be below 25.

Unfortunately I know this probably won’t apply much to AU market. Or maybe it still might. You know, US volatility extremes may still have some effect/result on AU market extremes etc…

Oh and when I say I had some success, I mean it mainly helped me reduce my drawdown during those ugly periods, it did not necessarily help make bucket loads of cash. Still, sitting it out and not losing money is sometimes the same thing as making money.

I did a quick search and it looks like there may have been something called the A-VIX but I cannot see it as a symbol in the Norgate data set. Might be handy if you could find it?

How are you doing your momentum ranking?

[MichaelRodwell]

Cool. Thanks for the tip. Will investigate.

For the momentum ranking I have tried a few things:

– ROC / ATR Combo
– Multiple ROC / ATR Combo’s with varying weightings on shorter and longer lengths
– ADX Multiplier
– ADX filter (ie min ADX)
– RSI filter
– MACD

And a bunch of other combinations of the above. So far I’ve concluded, they all work reasonable well most of the time so I decided to simplify as much as possible.

For this system, I have found the ROC by its self to be the most effective and the simplest. Between ROC(95) and ROC(250) the results in terms of CAR and DD are consistent.

For the system above all I have is:

– Index Filter – simple MA
– Momentum Ranking – ROC
– Monthly Rotation
– 7 Positions

Right now I’m reintroducing the stock filter and doing some more testing to see if that makes a difference. In the latest system iteration I’ve been able to reduce the days wait till equity highs to around a year which I think is palatable for a monthly system.

[Anonymous]

Have you considered using the Clenow method of linear regression ranking for the momentum?

I calculate it twice within my code, once for the symbol, and then once against the reference index I am running the system on. For example, if running it on S&P500 then I calculate it on the $SPX. Then once both calculations are done, I subtract the result for the index from the result for the Symbol. My theory here is that the momentum ranking for the symbol needs to be higher than the general market momentum (over the same period) to take a position with that Symbol. I use this number/calculation for my ranking method. Obviously you only take positions ranked above zero, those with the highest number is the highest ranked and therefore take those positions first.

I pasted the code below….

You can change the lookback timeframe days depending on what works for you in your system. I find on my shorter term, swing based systems when I’m holding anywhere from a week to a month or two, then about 50 days works. For longer term stuff where you might be holding for 6 to 12 months then higher numbers like 200 or 250 work well.

In this regard you should be careful though, because if you try and optimize too much the lookback days just keeps suggesting higher and higher values because it starts to favour anomalies that had really long bull runs in the last decade like NVDA. Actually I have NVDA, MSFT, MU and AMD in my EXCLUDE list when I am doing optimisation specifically for this reason, such that the settings I end up being happy with are not too heavily skewed by these stocks showing brilliant gains in the last decade because of cryptomania. Usually having these few missing from backtesting/optimisation means I might be (for example) showing 5000 fills in the last 15 years and then when I add them back in I get perhaps 5100 fills and the results look a little better. This is much better (in my opinion) than falling into an optimisation trap where you instead have a 2000 fill system which had plenty of trades on the crypto related symbols and when you exclude them your results really tank and you only have 1500 fills left.

//Symbol Momentum & Ranking
IndexCodeC = ParamStr(“Reference Index:”,”$SPX”);
IndexC = Foreign(IndexCodeC,”C”);

LookbackTimeframe = Param(“Ranking Lookback (Days)”,50,2,300,1);

LogOfPrice = ln( C );
ExponentialSlope = LinRegSlope( LogOfPrice, LookbackTimeframe );
AnnualizedSlope = 100 * ( ( exp( ExponentialSlope ) ) ^ 250 – 1 );
R2 = Correlation( LogOfPrice, BarIndex(), LookbackTimeframe ) ^ 2;
TickerRank = AnnualizedSlope * R2;

IndexCLogOfPrice = ln( IndexC );
IndexCExponentialSlope = LinRegSlope( IndexCLogOfPrice, LookbackTimeframe );
IndexCAnnualizedSlope = 100 * ( ( exp( IndexCExponentialSlope ) ) ^ 250 – 1 );
IndexCR2 = Correlation( IndexCLogOfPrice, BarIndex(), LookbackTimeframe ) ^ 2;
IndexCRank = IndexCAnnualizedSlope * IndexCR2;

RANK = TickerRank – IndexCRank;

[MichaelRodwell]

Thanks Matt. I did start with Clenow’s method for my first rotational system but haven’t revisited. I plan to launch my ASX system in Feb so this will give me some time to test this out.

Thanks for sharing the code!

Will report back with findings.

[MichaelRodwell]

Buy the Bulls Monthly Momentum

=> May: -7.79
=> June: +5.49
=> July: -0.54
=> August: -0.18
=> September: -1.72%
=> October: +3.46%
=> November: + 0.46

=> December: + 9.8%

YTD: +8.3%

Good end to the year! Not sure how I compared with the index as I’m out of town and don’t have any charts etc.

My goal in 2019 was to launch my system and take control of my trading 100%. I own these results and am very happy, especially after taking the month 1 slap in the face.

Happy new year all!

[MichaelRodwell]

Hey team,

I’m interested to hear if anyone does any testing around open profit or if that is even possible.

Right now about 18% of my total NLV is open profit.

It makes me feel a little nervous and I’m sure I’ll give some back but I’m wondering if there is a point in to take some off table AND if this can be tested.

Eg, if this got to say 50% open would there be any value in tighten up some stops, closing out positions and shutting up shop for the month?

[GlenPeake]

Hey Mike,

In short, I haven’t gone down that path of taking a bit of profit off the table…..or testing it….. I’d probably just hurt CAR and increase MDD…..I just trade the system as the signals are generated…… that way whatever the profit/loss result is, you can walk away knowing you made a successful trade sticking to your system etc……

Besides, if you were to take some profit off the table, you know what’s going to happen, don’t you???? The stock you partially sell is gonna come out with some ridiculous announcement, which sees the stock jump 20-30% higher and you’ll be kicking yourself, saying ‘why didn’t I just follow my system’…. I’ve spent countless hours testing it and crunching it North South East West etc….

I guess there’s a few sayings that float around regarding ‘profit’ etc…. i.e. the classic is…

….”No-one ever made a loss taking a profit”….

HOWEVER!!!!

….”No-one ever made a BIG profit by taking small profits”…….

If your nervous, no need to check your Monthly system on a Daily basis…… TELSA will still be there at the end of the month

Cheers
Glen

[Nick Radge]

Stop thinking about the money.

Think process.

[Anonymous]

I’m still struggling with the “stop thinking about the money” part. I think I am getting there and as time goes by and all the number of trades mount up I can see my edge playing out and I CAN see the bigger picture. I’m making progress but I don’t think I’m there yet.

So I am totally on board with holding the line and keeping the process on your current system. Mike I would keep the process as it is and follow what you have built, backtested, and you know what works.

That said, if you are starting to have these feelings of what you have stated about taking open trade profits off the table, firstly this generally only enters my mind when the market is up, so I guess it can be a good problem to have, but don’t overthink it. Second, my way through it in these cases is that if my mind is working against me in this kind of way and the juices are flowing on taking an alternative route, then take that energy and start building a new system. Put the idea into a new system of some sort, somehow, and see if it holds water. Backtest it, bend it, optimise it, do whatever you like to see if your idea is going in the right direction, for the long term. Hell, you may come up with something great! If you do, perfect! You can then add that new system to your arsenal, and start allocating some funds to it over time. This will satisfy your desire to have a “taking money off the table when everything is up” type of system and still keep your current, solid system running exactly as it should without naughtily poking your finger in and screwing it up.

Anyway that’s how I would probably tackle it. Nothing wrong with a nice little arsenal of systems working in different ways to satisfy different types of market conditions. In fact, most would say it is absolutely necessary to have multiple systems to help manage risk and help make money in all/most conditions as they present themselves.

[MichaelRodwell]

Not close to doing anything stupid yet Thank you Nick – back to sitting on my hands

[MichaelRodwell]

Glen – yeah, I realised this is a damned if I do or damned if I don’t proposition. I will be sticking to the system, just wondering if this is something that could add value. Many of us go to cash in a strong down market, so may be there is something in taking profits in a strong up market.

Matt – good idea re: focusing that energy elsewhere.

Thanks for the therapy session guys.

Regular programming has recommenced.

[MichaelRodwell]

Buy the Bulls Monthly Momentum:

What a difference a few days can make.

V Good month became a good month.

Ma19 -7.79
Ju19 +5.49
Ju19 -0.54
Au19 -0.18
Se19 -1.72
Oc19 +3.46
No19 +0.46
De19 +7.48

Ja20 +3.46

Total: +9.69

All time equity highs followed by a slap in the face.

Some good holds.

AMD +20%
TSLA +45%

Go live for my ASX system will be delayed because account issues with IB.

Getting ready for a turbulent Feb.

[MichaelRodwell]

Anyone trade the 1999 tech bubble?

Is TSLA behaving like its 1999?

[JulianCohen]

Just mentally prepare yourself for the drop that might not happen….

[MichaelRodwell]

Thanks Julian. Admittedly my anchor point is rising but I expect to give a lot back. The best I could hope for is some sideways action and a regulation rotation out of the position.

[Author]

Posts