Who remembers the 60s ?

[ScottMcNab]

I was looking through my options for upcoming data subscription when realised I must have selected Diamond when we were switched over. No point having virgin data sitting there so I started running some tests.

I really only have 6 systems …2 MRV, 2 rotational Momo (EOM) and 2 trend following systems (enter established trend…trailing stop). I moved them all to the SPX and broke the test into decades (60-69, 70-79 etc) and found some interesting results.
1. MRV (10×10) were the stars…consistent returns (even thru 70’s in US downturn) with reasonable maxDD
2. rotational momo (10×10) great 1995-current but awful cagr in 60s and 70’s.. first consideration is that my rotn momo systems are at fault. If anyone has this data and could run their system on SPX I would be interested in their results (I turned off index and price filters). Any theories on what else might have caused this?
3. the TF systems (20×5) were surprisingly resilient…kept chugging through the decades producing returns around 10% (does better on XAO and RUA but no data back to 1960)…these TF systems (hold about 6 months on average) would also have collected dividends along the way
4. While the performance for the MRV dropped in US in last decade, on XAO there was no drop off….take home message is to diversify across markets too I guess

[ScottMcNab]

[ScottMcNab]

[ScottMcNab]

Any other patterns in there or take away messages anyone can see? I’m not looking at actual figures (as none of systems trade on SPX) but rather the patterns. First area of debate may be whether it’s worth looking at data that far back at all?

[ZachSwann]

Could it be that you system punishes volatility and the ranks themselves are not that high?

[Nick Radge]

Nice work Scott. The main issues back then was that stocks were probably less liquid than today so some of the filters need to be adjusted – as you’ve rightly done.

Also, pricing wasn’t decimal back then and bid/offer spreads were probably considerably wider.

[ScottMcNab]

Is this in regards to the momo rotn Zac ? System 1 does indeed try and avoid volatile stocks …I did have a system I was playing with a few months back that sought out volatile stocks using HV>x ..I will dig that out and give it a run.

With regard to ranks, the idea here is to have a minimum hurdle for the rank ?

[ScottMcNab]

Thanks Nick. I imagine trading costs were higher too. I used the current IB values but maybe I should do a few runs with commission at 0.12-0.15 to be more realistic…would hurt the MRV most.

[ScottMcNab]

The rotn EOM system I have that prefences HV stocks was even worse for 60-70s…anyone else got diamond sub?

[ScottMcNab]

Rotn results were due to sloppiness in backtest…if reduce backtest eq to level should be for 60’s and/or use % for commission(0.2%) rather than cents/share then all fine…

[Author]

Posts