Could you win this game?

Tuesday, Jan 09 2018 by
Could you win this game

Suppose I proposed to you that instead of picking stocks this year, you could have a lucky dip and pick a ball from one of two hats, Hat A or Hat B. Each hat is full of several balls which are coloured either green or red and have a percentage printed on them, where green balls have positive numbers and red balls have negative numbers. The result of the draw is equivalent to your investing performance for the year.

First things first, you need to choose your hat. To help you decide I tell you that 71% of the balls in Hat A are green, whereas only 46% of the balls in Hat B are green. However the big attraction of Hat B is that there is almost a 7% chance of picking a green ball with a value greater than 100%, whereas in Hat A there is a only a 3.5% chance of picking a green ball with a value greater than 100%.

Still undecided?

How about I show you the 5% - 95% (5% lowest to 5% highest) distribution of values printed on the balls for each of the hats.



You would choose Hat A, right? Although if you are lucky enough to pick a high value winner it will be lower return than for Hat B, more of the balls have positive numbers and the loss given a negative number is significantly reduced. 

So why not do the same when picking stocks? 

As you may have guessed, the distributions shown above are not for balls in a hat but actually the performance of stocks in 2017. Hat A is the distribution of performance for stocks from within our top decile of StockRanks in 2017, that is the good, cheap improving stocks with the highest exposure to Quality, Value and Momentum, whereas Hat B is the distribution of returns for the bottom decile. 

As previously mentioned, you could have been lucky enough to pick a great winner from the bottom decile of StockRanks - congratulations to those who picked Falcon Oil & Gas or Oxford Biomedica - but the skew is certainly against you. 

Choosing a top ranked stock from the universe wouldn’t have guaranteed a positive return either, as many as 29% were losers in 2017…

Unlock this article instantly by logging into your account

Don’t have an account? Register for free and we’ll get out your way


As per our Terms of Use, Stockopedia is a financial news & data site, discussion forum and content aggregator. Our site should be used for educational & informational purposes only. We do not provide investment advice, recommendations or views as to whether an investment or strategy is suited to the investment needs of a specific individual. You should make your own decisions and seek independent professional advice before doing so. The author may own shares in any companies discussed, all opinions are his/her own & are general/impersonal. Remember: Shares can go down as well as up. Past performance is not a guide to future performance & investors may not get back the amount invested.

Do you like this Post?
48 thumbs up
0 thumbs down
Share this post with friends

28 Comments on this Article show/hide all

ed_miller 9th Jan '18 9 of 28

In reply to post #295178

Yes - You just enter the transaction details! - A bit of work but for that you get feedback and suggestions from highly successful traders like PhilH (I recommend you take a look at the rules for his fantasy portfolios here on Stockopedia, btw.)

(I will reply to your reply to me in the NAPS 2018 Performance thread soon, Natalie.)

| Link | Share
ed_miller 9th Jan '18 10 of 28

In reply to post #295178

Also - and this is just a suggestion, up to you what you do, obviously - but I think there is a case for holding 'forever' large, mature companies that consistently generate high returns and margins from stable revenue, especially sticky, recurring revenue, so long as you bought it when cheap. Even then, you'd need to monitor it at least every six months to check nothing in the company's operations or markets has changed for the worse. A 'hold forever' strategy would be particularly dangerous for cyclicals or small-caps or anything bought at a high valuation, for example.


| Link | Share | 1 reply
pippasfan 9th Jan '18 11 of 28

In reply to post #295218

That's a good point, but I just love the idea I'm still winning the game, even though Ive taken my money off the table. One other thing, I do like small cap companies, with strong balance sheets; in this instance they would be bigger balls inside the hat, for your choice

| Link | Share | 1 reply
ed_miller 9th Jan '18 12 of 28

In reply to post #295298

Cool, and that means you're still in the game if fundamentals are still positive and the price action (SP momentum) still supportive, but that doesn't mean you have to hold stuff indefinitely even if it is illiquid but has a strong balance sheet - if a good small-cap issues a profit warning that you judge is due to a structural change in its market (e.g. a high-street retailer suffering increasing competition from online disruptors) then I'd suggest at least trying to get out as soon as possible (8am on the morning of the warning if you can - if that doesn't work for your life you might need to rely on diversification, quality, not over-paying and only entering a trade when the SP action is supportive, not when it's still falling).

And it's worth remembering what Peter Lynch said about cyclical companies: "Cyclicals are like blackjack: say in the game too long and it's bound to take back all your profit." If you want to be able to hold half your position indefinitely then I think you should avoid cyclicals (or poor quality; and don't over-pay)!

| Link | Share
Andrew L 9th Jan '18 13 of 28

In reply to post #295118

Sorry you have so many thumbs down. Can't there be a stockopedia option for humour value? If there was I would give your post a thumbs up on that.

| Link | Share | 1 reply
pippasfan 9th Jan '18 14 of 28

In reply to post #295358

Many thanks haha I think

| Link | Share
andrewdb 9th Jan '18 15 of 28

In reply to post #295128

Running a fantasy fund is one thing.

However, we are supposed to be a little more systematic and use a method / procedure.

Surely it is more in the spirit of stocko to have a fantasy screen?

| Link | Share | 2 replies
PhilH 9th Jan '18 16 of 28

In reply to post #295478

My fantasy fund mirrors my personal investments with the exception that I didn't start with £1m.

It always amazes me why people 'listen' to the views of people on forums when they have no idea of their performance.

Everyone should have a fantasy fund and then we could more readily separate the wheat from the chaff.

Professional Services: Sunflower Counselling
| Link | Share | 1 reply
PhilH 9th Jan '18 17 of 28

In reply to post #295478

and what precludes anyone from using a method/procedure on a fantasy fund (whether there is real cash involved or not)?

Professional Services: Sunflower Counselling
| Link | Share
HHShah 9th Jan '18 18 of 28

I am new to Stockopedia.
I have seen the results you have achieved mentioned on the marketing literature for Stockopedia.
Where can I find the rules for your Fantasy Fund on the Stockopedia website?
Thank you

| Link | Share | 1 reply
PhilH 10th Jan '18 19 of 28

In reply to post #295523

In the fantasy fund description I provide links to the screens that I use.

I've recently attempted to rebalance the fund to more accurately represent the percentages i hold in my personal accounts.

Professional Services: Sunflower Counselling
| Link | Share
pippasfan 10th Jan '18 20 of 28

In reply to post #295498

This all sounds very interesting, something I’ve never even considered . However since I’ve been trading for nine years now, and some of the positions I hold, Iveheld nine years; how can that be replicated in a fantasy fund ? At the moment for instance the ftse100 is pretty high, and I’m looking round madly for short trades, can that be replicated inside a fantasy fund?

| Link | Share | 1 reply
pka 10th Jan '18 21 of 28

Hi Oliver,

Thanks for the interesting article. However it is unrealistic to expect many investors to select stocks from Hat B, which is the bottom decile of StockRanks. After all, if anyone goes to the trouble of using the StockRanks, why would they select from the bottom decile? I think a more realistic comparison would be to have a Hat C which contains all the stocks in the market that qualify for the StockRanks, prepare a graph showing the distribution of Hat C's returns, and compare that with the corresponding graph for Hat A.

| Link | Share
PhilH 10th Jan '18 22 of 28

In reply to post #295563

Just like everyone else, your fantasy fund has to start somewhere, so it'd be trades from today.

You can short shares in the fantasy fund but you can purchase short ETF's on indicies such as £SUK2.

I think doing anything 'madly' is not ideal.

Given you've been trading for nine years, can I ask what is your annualised return over that period?

Professional Services: Sunflower Counselling
| Link | Share | 1 reply
pippasfan 10th Jan '18 23 of 28

In reply to post #295743

I do everything madly Phil. Thanks for the advice, I’m taking up too much of your time here.
I normally make 40 to 50 % profits across my accounts , but I count up each March just because of the tax years. In my early trading years I was scraping maybe 15 %. As the great Robbie Burns says, trading without being able to go short is like fighting Mike Tyson with one hand tied behind your back, haha

| Link | Share
Nick Ray 10th Jan '18 24 of 28

Should I choose Hat A or Hat B?

Well, it depends what I am trying to optimise.

If I am trying to optimise the mean performance of my portfolio there is not enough information. You have chopped off the distribution plot at 5% and 95% so we cannot compute the mean accurately because we are missing 10% of the data. Now if the distribution were log-normal we would know that the outliers do not change the mean (or variance) very much. But for stock returns we know that not only do the outliers matter, but that they can be so significant as to completely overwhelm the data from the middle 90%. (The basic premise behind "The Black Swan" by Taleb.)

On the other hand, if I am trying to minimise a downside risk defined as "probability that portfolio returns less than some minimal acceptable return (MAR)" then I can read that data from the plot for most values of MAR. So we should choose whichever hat gives the lower probability reading for the specific MAR (since the probability on the graph gives the probability of failing to achieve the MAR.) We then see that for a MAR < 60% (ish) we should choose Hat A and for a MAR > 60% we should choose Hat B !

Incidentally I tried to quickly reproduce your graphs and I found that Hat A was always better than Hat B for all values of return. See plot below (which is created with a library CDF package so it has the axes transposed compared to your plots). This version would choose Hat A in all cases. (I don't know why I got such a different result. Did you filter by Market Cap? Also it's possible my data is skewed because Stocko does not distinguish between "no data" and 0%. Both return as "-". Maybe your plot is not a CDF and I have misunderstood.)


| Link | Share | 2 replies
pippasfan 10th Jan '18 25 of 28

In reply to post #295888

Excellent work, brilliant piece

| Link | Share
Nick Ray 14th Jan '18 26 of 28

In reply to post #295888

Well, I found my mistake and now (using my very limited data) I get the same crossover that Oliver has between low and high stock rank CDFs somewhere around the 60% mark. (Note: lowest curve is "best", or if you read horizontally, furthest to right is "best".) Of course that does raise the question of why the curves swap over. Perhaps something worthy of investigation by the Stocko team using their full database.

[Right-click and 'view image' to see a bigger version of the plot below]

| Link | Share
Sonia Danthy 18th May '18 27 of 28

Interesting! Now that I think of it - this should be quite intuitive. Thanks Oli!

| Link | Share
Blissgull 19th May '18 28 of 28

But could most people win at that game of drawing balls from a hat in the long term? It may not be as simple as it first appears.

The game reminds me of the famous Ralph Vince experiment. For those unfamiliar with it I will give a quick copy and paste description below


Known as the Ralph Vince experiment, he took 40 PhD. students and set them up to trade with a computer game. All 40 had doctorates, but Mr. Vince made sure that none of their doctorates involved any sort of background in mathematical statistics or trading.

The Experiment Challenge
In the experiment, they were given $1,000 and 100 trades, with a pre designed statistical 60% win rate. The rules were simple. When they won, they won the amount of money they risked. When they lost, they lost the amount of money they risked. So, after all 40 students had completed their 100 trades, how many do you think made money?

Shocking Results
Only 2 students out of 40 were able to make money, the other 38 all failed to succeed. Remember these were PhD. level people and they had 60% built in winning percentage. It means that they should have won 6 trades out of 10 on average. The odds were in their favor. Why did 95% lose money? Why did only 2 students out of 40 win? Does it make sense that these intelligent people lost?


The reason most lost - some completely wiping out their (imaginary) accounts - was that they tended to overbet. In the experiment described the optimal strategy would have been to bet 20% of your new equityl on each trade. Using that system the initial $1,000 would grow to about $7,500.

So having a good system for picking shares is only part of the story. Good money and risk management and an understanding of optimal position sizing is probably more important in the long term.

| Link | Share

Please subscribe to submit a comment

About Oliver Cooper

Oliver Cooper

I am an Analyst at Stockopedia - my job here is to get the most out of our fundamental data by exploring new ways to present and interpret it. My experience is predominantly based around Portfolio Construction & Risk; researching and implementing quantitative techniques to create better portfolios and make better decisions. more »

Stock Picking Tutorial Centre

Let’s get you setup so you get the most out of our service
Done, Let's add some stocks
Brilliant - You've created a folio! Now let's add some stocks to it.

  • Apple (AAPL)

  • Shell (RDSA)

  • Twitter (TWTR)

  • Volkswagon AG (VOK)

  • McDonalds (MCD)

  • Vodafone (VOD)

  • Barratt Homes (BDEV)

  • Microsoft (MSFT)

  • Tesco (TSCO)
Save and show me my analysis