ASX Small Caps Performance over the years? Backtesting trading system question

[Frankieplus]

Hey everyone.

I'm backtesting a systematic trading system that i want to use to trade the ASX Small Universe..
Its a trend following system

If I test this system for all of 1990-2000 ASX Small Caps the results are favourable.
From 2000 - 2020 the results are favourable up to 2010.

From 2010 until now (2020) the results are mediocre and probably not worth trading.

I'm not a veteran in the market and don't have history behind me to know if Small Caps were 'different back then'

So whats going on here? Have Small Caps just not been performing as well from 2010 - 2020 as they were in previous years?
Not as much 'bank for the buck' in the Small Caps index constituents?

Or does a trading system need to be tweaked over the years to 'fit' the market you are trying to trade?

Also, is there a way I can check the results of the S&P ASX Small index itself over the last 20 years so that I can see the performance of it?


Thanks..


-Frank

 

[rnr]

Hey everyone.

I'm backtesting a systematic trading system that i want to use to trade the ASX Small Universe..
Its a trend following system

If I test this system for all of 1990-2000 ASX Small Caps the results are favourable.
From 2000 - 2020 the results are favourable up to 2010.

From 2010 until now (2020) the results are mediocre and probably not worth trading.

I'm not a veteran in the market and don't have history behind me to know if Small Caps were 'different back then'

So whats going on here? Have Small Caps just not been performing as well from 2010 - 2020 as they were in previous years?
Not as much 'bank for the buck' in the Small Caps index constituents?

Or does a trading system need to be tweaked over the years to 'fit' the market you are trying to trade?

Also, is there a way I can check the results of the S&P ASX Small index itself over the last 20 years so that I can see the performance of it?


Thanks..


-Frank

Hi @Frankieplus,

The S&P/ASX Small Ords is known as the XSO and should be available from your data provider (e.g. Norgate Data).

Cheers, Rob

 

[Stockbailx]

Not to sure on the benefits of back testing the small caps. Not a real fan. I mean at best you could resolve the best sectors to trade, and I don't know how it works really but I'd be checking the financials of the stock and looking for compounded stock set to pounce.

 

[Frankieplus]

Not to sure on the benefits of back testing the small caps. Not a real fan. I mean at best you could resolve the best sectors to trade, and I don't know how it works really but I'd be checking the financials of the stock and looking for compounded stock set to pounce.

Interesting you say not sure about benefits of back testing small caps. Could you elaborate on that one? Is it because they are more volatile? I'm interested in your thoughts as I'm a beginner when it comes to backtesting anything

I am trying to use a trend following strategy to trade on a weekly timeframe. I chose small caps because I thought it would suite this strategy which basically buys breakouts and follows them up with a trailing stop.

The results are look great if I test it on All ASX stocks with prices from 0.50c to $10 but when I test on all asx small caps with no price filter the results are kind of all over the place a bit. Still ok over a 10 year timeframe but it has 3 or 4 losing years out of 10..


-Frank

 

[Frankieplus]

The S&P/ASX Small Ords is known as the XSO and should be available from your data provider (e.g. Norgate Data).

Ah of course. Thanks


-Frank

 

[tech/a]

Couple of things

(1) Survivorship
(2) Liquidity.

If your trading Smalls you'll get issues with both.
Haven't the time to train you.
But adjusting for liquidity say average of $500K traded a day
Only testing on 24 mths. of data maybe helpful.

From working with a number of quants I've come to the conclusion
that most readily available Testing software is at best rudimentary
and at worst misleading. Particularly at the hands of people like me
with very basic knowledge of programing and what Im looking for
or more to the point what I should be looking for.

(Although I've designed and tested many
profitably).

Frankly I trade a lot of smalls and there is BIG money to be made on capital invested.
Don't use a system trade discretionary.

 

[Frankieplus]

Couple of things

(1) Survivorship
(2) Liquidity.

If your trading Smalls you'll get issues with both.
Haven't the time to train you.
But adjusting for liquidity say average of $500K traded a day
Only testing on 24 mths. of data maybe helpful.

From working with a number of quants I've come to the conclusion
that most readily available Testing software is at best rudimentary
and at worst misleading. Particularly at the hands of people like me
with very basic knowledge of programing and what Im looking for
or more to the point what I should be looking for.

(Although I've designed and tested many
profitably).

Frankly I trade a lot of smalls and there is BIG money to be made on capital invested.
Don't use a system trade discretionary.

Good advice, thanks.

Since it's small growth companies I'm trying to target, If I trade all ASX stocks from $0.50 to $10 (but add $500k traded per day as you suggest) in you opinion would I be targeting mostly small growth companies anyway?

I know trading All ASX 0.5 to $10 isn't officially the small caps because there could be larger companies in the mix where their share price happens to be low..

Basically I see no way to get my strategy to give me the results I want targeting the constituents of the index I want to trade (XSO) so I'm trying to do a bit of a workaround by targeting the 0.50 to $10 range of all of the ASX instead.


-Frank

 

[Stockbailx]

I'm no pro at back-testing either. Being Volatile would have nothing doing. You would be finding stock from say $1.50 to confirm a trend. trading a weekly time frame, must make you confident, but to find trending stock would be ambitious but may be hard to do with small caps. So I still say if its a part of back-testing, confirm stock with good profit compound at all angles. To trade breakouts, you would want to be buying from the support level (trailing stops) and following the breakout out. to proceed with the trend...

 

[qldfrog]

Good advice, thanks.

Since it's small growth companies I'm trying to target, If I trade all ASX stocks from $0.50 to $10 (but add $500k traded per day as you suggest) in you opinion would I be targeting mostly small growth companies anyway?

I know trading All ASX 0.5 to $10 isn't officially the small caps because there could be larger companies in the mix where their share price happens to be low..

Basically I see no way to get my strategy to give me the results I want targeting the constituents of the index I want to trade (XSO) so I'm trying to do a bit of a workaround by targeting the 0.50 to $10 range of all of the ASX instead.


-Frank

Do not use SP for small cap determination.many below 20c.some major companies below $10
And then a 10$ sp 15/20y ago is probably a $15 now just with inflation..then as market crashes some will jump in out of your realm...

 

[divs4ever]

for small 'growth stocks ' maybe you need an extra parameter ( say a market cap. ... above $4mill perhaps , or positive EPS )

i went a lot of times for small caps that paid divs and sure i got burnt several times , but not as often as just trying for those likely to triple in price ( and hope i can sell them that day )

but maybe the strategy needs a couple of tweaks ( tweaks are fine when they work for YOU )

but sure back-test away , it sure beats leaking cash ( imo )

cheers

 

[Belli]

boring as bat poo EX20 is one way or other ETFs in that sphere.

ETPs

Exchange-traded products (ETPs) on ASX offer access to a sizeable market providing variety and choice across over 200 funds.

www2.asx.com.au

Throw in VISM for international is also a possibility.

Likely to do better than fiddling around thinking doing something beats doing nothing. Long run it don't.

 

[Frankieplus]

To trade breakouts, you would want to be buying from the support level (trailing stops) and following the breakout out. to proceed with the trend...

Yes thats what the strategy does. Buy at 20 week breakout and trail stop..

-Frank

 

[divs4ever]

Do not use SP for small cap determination.many below 20c.some major companies below $10
And then a 10$ sp 15/20y ago is probably a $15 now just with inflation..then as market crashes some will jump in out of your realm...

indeed ,

AMP and ORG are still large companies

i would re-suggest a market capitalization parameter say between $2m and $20 million ( but it is YOUR back-test , pick what suits you )

 

[Frankieplus]

Do not use SP for small cap determination.many below 20c.some major companies below $10
And then a 10$ sp 15/20y ago is probably a $15 now just with inflation..then as market crashes some will jump in out of your realm...

Cool thanks.. . Could you suggest basic filter I could use to find small caps growth using purely technicals? Other than volume and market cap there wouldn't really much else would there?

I know we probably need fundamentals thrown in the mix here to determine the growth but I will use Stock Doctor for this after the technical scan for the Small Caps find me the candidates.


-Franke

 

[Stockbailx]

Yes thats what the strategy does. Buy at 20 week breakout and trail stop..

That would be hard to Identify, good luck on finding that elusive stock. All the Best!

 

[qldfrog]

Market cap is probably all you can get easily...

 

[Frankieplus]

That would be hard to Identify, good luck on finding that elusive stock. All the Best!

That would be hard finding just 1 stock

It's a portfolio I'm building of 20 stocks.. Not just the one stock.. So any stock (hoping small cap) that breaks out of a 20 week range or high becomes part of the 20 stock portfolio, and then trailing stop.


-Frank

 

[divs4ever]

also maybe thinking of your system as a way to REDUCE bad outcomes , rather than find the unicorn , could be helpful ( all my 6 baggers were a total surprise to me , even now for a couple )

 

[Stockbailx]

Market cap is probably all you can get easily...

Would weekly Indicators be of any significant value. Use an Indicators with the correct appetite and settings and you might get yourself on to a good wicket. e.g. might be ATR, MACD, Trend trigger factor. See what you've got in your goodies bag...Try a monthly indication for clarity...

 

[Frankieplus]

Would weekly Indicators be of any significant value. Use an Indicators with the correct appetite and settings and you might get yourself on to a good wicket. e.g. might be ATR, MACD, Trend trigger factor. See what you've got in your goodies bag...

Could be an option actually. I like the idea of that. I'll see whats in the showbag and keep tinkering