How much history – Part II

How much history – Part II

You are currently viewing How much history – Part II

Is the seniority of the strategy a really good measure?

We will predict the future of a strategy

On a previous post I analyzed 2,139 audited strategies coming from Darwinex, they are called Darwins in fact. This is a sample big enough to give us some statistical value. And from here crunched numbers with simple Excel exercices to try to arrive at any conclusion that could make any sense.

But what is the purpose of all this mess? Well, that’s a tough question. I’m trying to measure my own Darwin I guess. Also it could be a way to filter the good Darwins from the fishy from the great, in case I was an investor, for instance.

Summary. The work done on my previous article

Raw data

As my own Darwin is very young, my concern is to forecast if it has any chance to succeed in the future or it will fade away as so many other ones. So the longevity is a key factor. I’m taking the age of the Darwins as an axis to project into the future.

And how many Darwins are we talking about from now to the future on the Darwinex environment? Numbers on the table below:

I filtered these Darwins by its Return on the last 2 years and got the table below:

What I have learnt so far

Here some conclusions I’m pretty sure of:

Just about 7% of the Darwins who make their first birthday can make a living

After the second birthday you can be quite confident on the forecast of a DarwinDarwins with >20% returns a year, year after year, are just a chimera

Better you read my previous article, right?

Give me a measure and I’ll move the World

Good. Let’s give it one more thought and take a measure. The purpose is to try to predict the quality of a Darwin from a simple parameter, just reading a value. That’s a measure, right?.

What is the only measure that we can extract from Darwinex?, aside from the atributes which almost by definition only give us a piece of information and are not intended to bring us the full picture. So, unfortunatelly, what is the only measure we’ve left from Darwinex? Yes D-Score. I’m not a fan of the D-Score, but it isn’t a bad measure either. Personally, I’d choose Profit Factor or Sharpe ratio, simply a matter of taste. However we can’t choose. We’ll stay with D-Score, at the end of the day it could be as good as any other.

So, back to the point. The question is if we could classify a Darwin just with the result of its D-Score (again, a measure as any other). Let’s turn the Excel machine on and this is what I’ve found:

I think it’s not bold to conclude just what the numbers say:

A Darwin under D-Score of 66 – 67 is not worthy

Doesn’t matter how much experience it has, the results are quite stable and uniform through the years..

And now… Prove me wrong!

You know, things in trading are never cristal clear. If D-Score wasTHE measure, that would mean that just reading the D-Score of a Darwin I could extract its Return. Well, let’s come back to Darwinex and test my statement:

I have filtered all the Darwins – a total of 3,054 Darwins – by the D-Score and from here I have measured its Return over the past 2 years; my very first table on Part I, remember?

This is what we get:

If my hypothesis regarding the D-Score was right I’d like to see both #Darwins columns the same. In other words, the % should be as close to 100% as possible. Rather, we see that the difference on the bad Darwins is around 15% – could be taken as acceptable – but the difference between the good ones are too big.

Heck! Why is D-Score not predicting the Return of the strategy?? Three options:

  1. D-Score sucks. That’s the easy one, if your measure doesn’t satisfy your expectations just change it. Nahh! I don’t buy that
  2. My conclusions suck. Of course that’s an option, but in this case I should investigate why do they suck, where am I wrong
  3. Lack of statistical significance. I think this is the more likely explanation. Notice that the worst correlation between D-Score and Returns is on the best segments which have a population of 141 and 15 members

Some last words

I know, this post is too long. But right now my Darwin EPG has a D-Score of 62. Oh my God, I’m doomed.

Before I could even think to give up, I’ll better remind that we can change the future, we have the power to rewrite history. Darwins are not static. In the meanwhile, just enjoy!

Leave a Reply