Forward Testing The Calibration Score

6th Sep '20 by HollandSmith

12 comments 824 reads

This was purposely written short to not waste your time.

About me: Skip
Introduction: Skip
Substance: I have scored mostly all FTSE companies with a home-brew “Calibration” score which I designed to show which stocks are most likely to outperform in the next 25-50 days. This slowly changes everyday, but attached is a static snapshot for the opening bell on Monday the 7th of September in order to publicly forward test the mechanics of the score.

Please download a copy of the dataset here and rejoin on the 11th October (25 trading days) for the results.

Questions you might have

Q: Is this like the momentum rank?
A: Kinda, its scored from one hundred been best to zero but doesn’t use any %52w high, MA on price or RS. It’s also not a rank but an actual score which in my personal opinion is superior.

Q: Have you timed this blogged output to get the very best results?
A: Nope. The average of all the scores is well below 50, so I may be foolish publishing this right now. But there’s always an excuse so if not now, then when?

Q: Why is a certain company not on the list?
A: The output is contains 1,361 stocks. I may have taken out a stock if its a low market cap, extremely volatile or little to non daily change to price.

Q: Can I ask you a question?
A: Sure!

Unlock with Email

Unlock with Google

Already have an account?
Login here

Name

Ticker

Sector

Score

Change

Braveheart Investment

BRH

Financial (Investments)

100

127.90%

Fusion Antibodies

FAB

Healthcare (Research)

90.00%

Omega Diagnostics

ODX

Healthcare (Supplies - Equipment)

82.41%

Solgold

SOLG

Materials (Metals - Mining)

44.62%

Genedrive

GDR

Healthcare (Services - Providers)

34.74%

Sareum Holdings

SAR

Healthcare (Research)

29.00%

Tekcapital

TEK

Technology (Services - Software)

100

26.67%

Wey Education

WEY

Household (Goods and Services)

100

17.08%

Yourgene Health

YGEN

Healthcare (Research)

100

14.82%

Evgen Pharma

EVG

Healthcare (Research)

14.58%

Hikma Pharmaceuticals

HIK

Healthcare (Pharmaceuticals)

9.28%

Anpario

ANP

Household (Consumable)

8.31%

Trans-siberian Gold

TSG

Materials (Metals - Mining)

4.31%

Plus500

PLUS

Financial (Investments)

4.17%

E-therapeutics

ETX

Healthcare (Research)

3.93%

Dp Poland

DPP

Household (Entertainment)

100

3.57%

Bh Macro

BHMG

Financial (Investments)

100

3.12%

Intelligent Ultrasound

MED

Industrial (Services)

1.80%

Dev Clever Holdings

DEV

Technology (Services - Software)

100

-0.56%

Domino's Pizza

DOM

Household (Entertainment)

-0.76%

Northamber

NAR

Household (Retail)

100

-1.69%

Naked Wines

WINE

Household (Food - Drug)

-1.90%

Lidco

LID

Healthcare (Supplies - Equipment)

100

-2.86%

Smart Metering Systems

SMS

Technology (Equipment)

100

-4.39%

Deepmatter

DMTR

Industrial (Services)

100

-6.00%

Volvere

VLE

Household (Consumable)

-10.97%

Byotrol

BYOT

Household (Goods and Services)

-16.67%

Condor Gold

CNR

Materials (Metals - Mining)

-16.92%

C4x Discovery Holdings

C4XD

Healthcare (Research)

-21.55%

Pires Investments

PIRI

Financial (Investments)

-29.67%

Deltex Medical

DEMG

Healthcare (Supplies - Equipment)

-36.50%

HollandSmith

18th Oct '20

Results - purposely written short again

The results below are everything that the code scored over 90.

Name	Ticker	Sector	Score	Change
Braveheart Investment	BRH	Financial (Investments)	100	127.90%
Fusion Antibodies	FAB	Healthcare (Research)	98	90.00%
Omega Diagnostics	ODX	Healthcare (Supplies - Equipment)	91	82.41%
Solgold	SOLG	Materials (Metals - Mining)	92	44.62%
Genedrive	GDR	Healthcare (Services - Providers)	93	34.74%
Sareum Holdings	SAR	Healthcare (Research)	98	29.00%
Tekcapital	TEK	Technology (Services - Software)	100	26.67%
Wey Education	WEY	Household (Goods and Services)	100	17.08%
Yourgene Health	YGEN	Healthcare (Research)	100	14.82%
Evgen Pharma	EVG	Healthcare (Research)	92	14.58%
Hikma Pharmaceuticals	HIK	Healthcare (Pharmaceuticals)	97	9.28%
Anpario	ANP	Household (Consumable)	95	8.31%
Trans-siberian Gold	TSG	Materials (Metals - Mining)	92	4.31%
Plus500	PLUS	Financial (Investments)	94	4.17%
E-therapeutics	ETX	Healthcare (Research)	95	3.93%
Dp Poland	DPP	Household (Entertainment)	100	3.57%
Bh Macro	BHMG	Financial (Investments)	100	3.12%
Intelligent Ultrasound	MED	Industrial (Services)	99	1.80%
Dev Clever Holdings	DEV	Technology (Services - Software)	100	-0.56%
Domino's Pizza	DOM	Household (Entertainment)	90	-0.76%
Northamber	NAR	Household (Retail)	100	-1.69%
Naked Wines	WINE	Household (Food - Drug)	92	-1.90%
Lidco	LID	Healthcare (Supplies - Equipment)	100	-2.86%
Smart Metering Systems	SMS	Technology (Equipment)	100	-4.39%
Deepmatter	DMTR	Industrial (Services)	100	-6.00%
Volvere	VLE	Household (Consumable)	92	-10.97%
Byotrol	BYOT	Household (Goods and Services)	93	-16.67%
Condor Gold	CNR	Materials (Metals - Mining)	99	-16.92%
C4x Discovery Holdings	C4XD	Healthcare (Research)	92	-21.55%
Pires Investments	PIRI	Financial (Investments)	93	-29.67%
Deltex Medical	DEMG	Healthcare (Supplies - Equipment)	96	-36.50%

Overall 12% gain over the time period - which is in line with the codes historical track record of doubling annually.

I wish to publish again, but I will wait until the output has more substantial recognised names.

Thanks
Dan

Gromley

19th Oct '20

Well I'm finding it interesting anyway Dan!

12% in 5 weeks is certainly not to be sniffed at!

The pareto principle is on clear display here - you top 20% (6) delivered 68% gain over the period and the rest lost 1.6%. Is that typical from the earlier runs you have done?

Although that might make it seem like you have been "lucky" - in fact I find that a lot of structured portfolios have similar characteristics - oftentimes it is about fishing in the right pool to be able to find a small number of those super performers. (Over a longer time frame Ed Croft's NAPS has quite a similar spread).

The purist in me always wants to try to work out how to hone in on just the top 20% performance (in advance obvs!) but that is probably a fools errand. Applying "systems" in a total different field I have often found that apparently logical fine tuning can easily produce massively adverse results.

Anyway - given that this is essentially a trading strategy, have you thought about whether stop-losses might improve results? Or maybe my last paragraph applies?

20th Oct '20

Great to see your name appear and thanks for the comment!

When looking in hindsight to an ordered returns% list, some will instantly conclude the top returns are down to “luck”, and if these were not included, the overall return would actually be a loss. But I’m glad your trained eye notices many these systematic structured portfolios do share similar characteristics in their return distributions.

But I have to admit there was a bit of luck in play with this posting as I usually run the code daily in order to see what’s working/ not working and act accordingly. As I had published a static list, I was banking on the in-built risk/reward to play itself out, which I believe it did to an extent.

On average, each stock in the FTSE market has spent 10% of their time in a score 90+, performing around 600% vs. the under 90s. Does sound impressive, but when the under 90s spend the majority of time losing or trading sideways, the overall performance % doesn’t quite match.

Looking at the published static list compared to what I actually traded is useful and is going to be something I will continue in the future. Sparing the details, the classic run your winners, cut the losers is something I practice and preach. I know this doesn’t quite ‘hone in’ on finding the top 20% in advance, but does go a long way to make the overall performance greater.

The main reason why I use this code is because it forces me to focus my attention on the charts which I know are showing all the correct signs. It’s removed a lot of psychology mind play and allows me to actively see that I am constantly replacing capital - quickly cutting losses and taking healthy profits - whilst promptly opening positions in proven statistical winners.

Anyways, need to have a good think about how I can show it off in a better light as it potentially could help a lot of people.

Thanks for the comment
Regards, Dan

dfs12

Hi Dan

Thanks for such an interesting and original post. I love this sort of thing and put my own rankings together too. Not sure they are worth public consumption though.

I'm interested to know a little more about the factors you use to create your ranking. Obviously I don't expect you to spell it out in full detail (unless you want to!) . Is it made up of mainly technical signals e.g. RSI values, MACD etc. Does it have any fundamental analysis e.g. EPS growth, PE values etc. And... does it factor in any positive newsflow? (obviously something you have a certain reputation for!)

Many thanks

Rusty2

It is a very good result but 12% in 5 weeks is a lot more than 100% per annum. I did have a look at Braveheart Investment (LON:BRH) on Stockopedia but could not see why it scored 100, could you enlighten me how your scores are arrived at?

Most of the top gainers have fallen back in the last week, so it looks like 25 days is probably the optimum time going by this result. I think there is some luck in all these type of selections. I would also be interested is what some of the high gainers were over that period, such as Novacyt SA (LON:NCYT) which was up over 100%.

I could not get the download to work apart from the first 100.

PhilHofGrowYourDough

Hi Dan,

Firstly apologies for missing this thread first time around.

I want to say thank you for being so generous as to share some details of the approach you are exploring. I want to ask some questions and it could that you want to keep some details close to your chest but if we don't ask we'll never know so here goes ...

1) have you been able to back test this approach?

2) Are you using this approach and backing it with funds?

3) Does the approach draw upon any particular recognised investment strategies? If so, are you able to be specific or give us a steer? For example it seems reading between the lines the share price movement is significant and extreme volatility excludes some selections.

4) Is the approach technical only or are you using fundamental or fundamental dervied factors such as QualityRank to build you Cailbration Score?

5) It would be interesting to see the performance of the other deciles. For example does the lowest decile provide a viable list of shorts?

6) what do you want from the Stocko community?

There are many here who can provide a friendly critical eye but it can be challenging to do that if we have limited visibility of the constituents and construction of the Calibration Score but I could understand a reluctance to expose that.

Once again many thanks for your generoursity.

Phil

21st Oct '20

Hi,

Based solely on price willing to be paid for an asset at a given time. I used to work in marketing agencies analysing customer behaviours to try improve the targeting and timing of communications. Despite being completely different subjects, it’s remarkable how some techniques complement each field.

News is not included in the score, but it’s quite amazing how often positive RNSs appear for charts been scored over 90. It’s something that I have gotten used to, but I probably need to be more grateful for.

Cheers!

Hi Herbie,

You’re correct to say 12% in 5 weeks is a lot more that 100% per annum. One could also consider some stops hitting the weaker ones, and been more selective with picks could increase this again. Not to forget the compounding effect of only holding on average for 25 days and recycling the full proceeds into a fresh set ups. But, I always like to focus on process than building “twitter” hype.

I will label up the Braveheart Investment (LON:BRH) chart for you the way I see it.

When it comes to Novacyt SA (LON:NCYT), I actually removed it from the output as it was showing signs of extremely volatility and was going to skew the results. With hindsight, these should of been left in as it would have had a great impact on returns.

Cheers!

Chart for Braveheart Investment (LON:BRH)

5f8fd69ecb1e9braveheart-investment-21-10

How neat would it be to publish a chart to the discussion board when drawing! Could be an embedded iframe which you could then fork to copy across drawn lines and comments. Just a thought. :)

Hi Phil,

1) have you been able to back test this approach?

I believe so. I have made an oscillating back tester which compares how well a charts higher score does against its own lower score. On average, 90+ scores are 6* better. 95+ are 43* better.

2) Are you using this approach and backing it with funds?

Yes, funds and trusts. Maybe that would be a better to public forward test going forward!

3) Does the approach draw upon any particular recognised investment strategies? If so, are you able to be specific or give us a steer? For example it seems reading between the lines the share price movement is significant and extreme volatility excludes some selections.

I only really removed extreme volatility from the output as I didn’t want an outlier to spoil the forward test. But, it’s fair to say that there is an element to the code which prefers declining/low volatility. Looking at the chart I produced, my approach is nothing Earth shattering, just all the classic goodness. Trending upwards after a pull back, filled with higher highs and higher lows. The only thing which I have really done is found a way to proactively score it.

4) Is the approach technical only or are you using fundamental or fundamental dervied factors such as QualityRank to build you Cailbration Score?

This score focuses my attention to the set up at the ‘correct’ time, which gives time to analyse the assets fundamentals behind the chart. I always like seeing high stockopedia ranks, a good story, or an endorsement in the discussions, but my main driver has to be my own score as it allows me to press the necessary buttons when the time arrives.

5) It would be interesting to see the performance of the other deciles. For example does the lowest decile provide a viable list of shorts?

I’ve been focused on the long side that I have neglected this. If I do a forward test with funds I will include this!

6) what do you want from the Stocko community?

That's such a humane question which deserves the most honest answer. I think it would be amazing to work on Stockopedias offering to make the best set of tools possible for the active investor, so hopefully posts like this will catch their attention.

Many thanks for your reply and extra info. Very interested to hear about your approach and background.

HumourMe

22nd Oct '20

Well done for systemising the various components. I've not managed to do so yet but still aspire to (I have the individual components). It is clear that in isolation nothing works and that the whole system needs various criteria.

As I understand the chart you have:

A long term trend proxy (rising 200d ma)

A pullback proxy (distance from 50d ma or touching 200d ma?)

Volatility contraction (how do you measure this?)

A breakout trigger (price > 50d ma peak) after rising higher lows (above 50d ma)

Do you do anything with volume?

What is your universe? Anything you rule out immediately? (apologies if you have already covered this).

How did you create the composite score? Did you use a multivariate technique to weight the inputs or just sum, product or average ranks. Did you establish any absolute thresholds or criteria required?

Forward Testing The Calibration Score

Unlock the rest of this article with a 14 day trial

About the Author

HollandSmith

12 comments

Unlock the metrics for all these stocks and more...

Unlock the metrics for all these stocks and more...