One Year Results For The Mechanical Bull Portfolio

30th May '14 by Mechanical Bull

11 comments 21456 reads

Please extend a warm welcome to the writings of the "Mechanical Bull" - a UK based private investor investing his portfolio with a purely systematic approach based upon Stockopedia data including the StockRanks and Screens.

For an introduction to his method please do read his excellent introductory post "The Mechanical Bull Method". We will be syndicating occasional articles from his blog to Stockopedia.

A First Birthday

On 24 May 2013, the Mechanical Bull portfolio was born. It wasn't called that in the beginning. It was just an idea that I had been thinking about for a while before I decided to put into action. I invested a hypothetical £30K to see where it would lead. It was a couple of months later when I gave it the present moniker.

So where are we a year later? Well according to stock prices on Google Finance, the MB portfolio is £42,779, which is an annual gain of 42.6 per cent. While this may seem impressive, it is important to see performance in perspective.

Comparative Analysis

First, we should look at the broader market. The FTSE 100 has been pretty flat over the past year. The FTSE 250 has done somewhat better overall with a rise of about 9 per cent, although it was up by a lot more earlier in 2014 with a big pull back in the last few months.

It is also useful to compare performance against other investment strategies. Stockopedia's "Screen of screens" gained by 25.6 per cent

The MB portfolio would have come 7th equal out of Stockopedia's 65 guru (long) strategies. However, three of those ahead have two or fewer stocks, and so the MB portfolio would have faired even better if these non-diversified strategies were excluded.

Although the MB portfolio has gone sideways for the past 3 months, it has actually held up quite well compared to other strategies and the broader market. Indeed, the MB strategy outperformed all these comparators on 3 month, 6 month and 1 year timescales as this table shows:

Table 1: Portfolio performance up to 25 May 2014

It's Not All Good News

It all seems pretty positive so far. However, a closer look at the current state of the…

Unlock with Email

Unlock with Google

Already have an account?
Login here

Edward Croft

30th May '14

Mechanical Bull. Your study validates the results we've been seeing. The top 1% StockRank stocks have returned about 40% over the 13 months or so. We'll be releasing this data soon. The top 10% of StockRank stocks have returned about 35%. It's been quite eye-opening really.

Of course - past performance is no indicator of future results. I'd be v. surprised if the next 12 months were the same - the stock market is rather unpredictable year to year.

Blissgull

31st May '14

Those are impressive results.

Can you clarify what rules have been used for selling shares in the mechanical bull? Are they sold when the stockrank+guru screen formula falls below a threshold or when they no longer appear in the screen of screens etc?

Is the buying formula just to take the top 15 shares by stockrank and number of screens?

mpat89

Thanks for posting this. Very useful. I have been using Stock Rank > 90 as one of my mechanical filters but your method is surely an enhancement to that. One thing that strikes me is that you have no rules regarding sectors. Would you not be concerned about becoming exposed to too much sector specific risk?

If we look at the current screen, there are 7 stocks with a spread of less than 5%, 5 of which are in Industrials. Any comments on mitigating that risk?

pooledaniel

@ Bliss - the trading rules are set out on the blog.

See http://bullmechanic.blogspot.co.uk/2013/12/the-mechanical-bull-method.html

Mechanical Bull

Blissgull - The sell trigger is when Stockrank+Screen of Screen count (MB score) falls below 90. The proceeds are then used to buy the stock with the highest MB score stock not already in the portfolio. This is a fairly arbitrary rule I have set just as way of being able to rotate the portfolio and maintain exposure to higher scoring stocks.

I have not done much detailed research into the effectiveness of this strategy. However, as part of my annual review, I made a comparison of the MB methodology as I've described with the performance of a portfolio of the 15 original MB stocks over the last year (i.e without any rotation). I think it is good idea to keep an eye on the rear view mirror to see how sold stocks have performed.

As it turned out, the performance was almost identical. So, maybe the current strategy is too cute and I would better off holding stocks for longer. Certainly it chimes with general warnings against over-trading. This is a theme I was going to write about a bit more in my next blog post.

mpat89 - You are right that the MB strategy over-weights certain sectors. Three of the current 15 MB stocks (Staffline, Harvey Nash, Matchtech) are in the recruitment sector. However, all three have all been stand out performers. So, I am aware of this issue, but I have not been excessively worried about it.

It would be interesting to go back and see how a more sector diversified strategy would have performed. That's given me a good idea for a future post. Thanks!

MB - we do have plans to build out a rebalancing/portfolio building module that would sit on top of the screener. i.e. one could filter screening/ranking results to ensure sound sector diversification, set up periodic rebalancing schedules (quarterly, annual), liquidity filters (filter out wide spreads etc) and also desired weighting (equal weighting, value weighting, mkt cap weighting).

Indications are that enforcing max sector exposures reduces volatility and risk (e.g. banks in 2008) but at the expense of some upside performance. It's all a bit of a balance.

Anyone reading this should join the StockRanks webinar on Wednesday - I'll be showing the first year results of the StockRanks plus a few thoughts about how best to use them. You can sign up here.

Will this webinar be recorded for us ordinary folks with day jobs?

PhilHofGrowYourDough

Hi MB,

Thanks for your blog and posts.

Have you considered that the European Screen of Screens has a 1 year return of 47% whilst the Eurofirst 100 has returned ~ 12%. The challenge with the European stocks is that you are going to get the more obscure exchanges, such as Poland and Turkey, plus you might also be uncomfortable about particular countries e.g. Greece, Spain, etc. That said of the top 20 stocks in the European SoS I have have owned £NOLA_B (Stockholm), Dart (LON:DTG) (London), ST Ives (LON:SIV) (London) & £BAKKA (Oslo) and I still hold Veidekke ASA (OSL:VEI) (Oslo), Cegid SA (EPA:CGD) (Paris) & Salmar ASA (OSL:SALM) (Oslo).

Also I think one of the issues with using the SoS as a starting point is that you run the risk of missing the biggest gains. I've found that my best returns have come from stocks that I've screened that appear on few screens then as they've developed they creep into the Screen of Screens. In other words the SoS sort of validates my original selection and rather than jumping into the share at that point I'm monitoring them looking for a change in momentum to signal an exit.

That said the mechanical strategy you describe is straightforward to follow. It'll be interesting to see how it continues to perform.

Best of luck
Phil

Hi mpat89 - yes of course - we'll record it and post it on Youtube.

1st Jun '14

PhilH

I have not looked at European stocks, partly because of the subscription costs, but mainly because my approach is a work in progress and for the moment, UK stocks are a big enough universe in which to test out these ideas

I tend to agree with your view on limitations on the SoS. Indeed, I think that screening in general has fundamental limitations. Basically, the more filters you apply, the fewer candidates you are left with so there is a practical limit on how many factors you can include. Also, a stock only has to fail on one criteria and it will disappear from the list. Screening is a blunt instrument.

The idea of StockRanks is much more accommodating. As there is no filtering, there is no practical limit on the number of measures that can be incorporated. The only question is which factors correlate with future performance and how these should be weighted.

The other thing I like is that StockRanks are a relative measure of how well a stock is fairing against all others. One of the problems with many of Stockpedia's guru screens is that they fluctuate over time and so some only have a 1 or 2 stocks, or even none. I think it is always better to think in relative rather than absolute terms and this is the essential difference between ranking and screening.

In future, I might look at ditching the screening element of the MB strategy completely. If Ed's comments about the performance of the top 1% StockRanks are anything to go by (i.e. 40% over the last year), then it may be that the SoS element is not actually adding that much.

MB

One year results for the Mechanical Bull Portfolio

Unlock the rest of this article with a 14 day trial

About the Author

Mechanical Bull

11 comments

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

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