Hi,
I’m wondering if anyone has suggested this already. It seems one of the difficulties evaluating systems is that, the longer the system has been around, the more capital it has to trade. Assuming the system is profitable, that is… This can affect the size of the returns and lead to a biased view of the system.
Is there an advantage to making the amount of trading capital available static? I think this would really make it easier for subscribers to figure out how to trade the system in their own accounts. The growth of the C2 account would then be tracked against the static account size. If each trade were calculated as a percentage of the original capital, we could simulate the growth of reinvesting gains by scaling the trades up to track the increase/decrease in the size of the account. So we’d be able to track 2 types of growth, that with reinvesting gains, and that without.
Cheers,
-lcg
Loren,
Yes, this has been suggested already. I (among others?) suggested it in my posts in the thread “Subscribers need protection from bad systems” of the Suggestions forum on (I think) 01/22/06. I had essentially the same arguments as you mention above. There are perhaps some practical complications (see the discussion in the earlier thread) but I still think that a fixed account size would be better.
Jules
When C2 started nearly four years ago (!!), we did use static account sizes. (Effectively we normalized trade sizes, so that all trades used the same amount of capital.)
We abandoned that concept fairly early in our existence because tons of people complained about it and very few people understood the reasons why we did it.
Since we already tried it and it was a disaster, I think it’s unlikely we will re-try it in exactly the same way. Perhaps a better way to act on your suggestion is to add a feature which allows system evaluators to apply a kind of “normalization” to each trade, and to see the equity curve “as if” the trade size had been constant. That’s not as simple as it sounds, but it may be a good feature to add to the To Do List.
Maybe an equity curve logarithmic option button can be added. This feature would basically convert the equity curve into log’ chart. This should be very straight forward to implement.
It is definitely a good feature to be on the list. As more systems start to show large gains, you will get more similar requests.
Ah, that explains a lot. And I understand that people don’t like it. But I would certainly appreciate such a feature very much.
Jules
There is a simple solution to the problem. The subscriber has to watch what % of the whole account the vendor is trading.
To keep it static, a successful vendor should keep increase the played capital to stay on the same static level. Some subscribers complained about this, that it is basicly pyramiding. It just shows you can not statisfy everybody.
But if a succesful vendor keep the same capital risked all the time, his static level actually will decrease over time.
The point is, with a little math subscribers can do both ways for themselves based on the data already provided.
Pedro,
This would not require ‘a little math’ but a tedious downloading of all trades for each system. Therefore I would appreciate it if C2 will do the math. Moreover, the present results are somewhat misleading for people who are not aware of the compounding effect. This will probably include many newbies.
Jules
This is true, Pedro, but try setting that up with autotrade. Every trade would require the subscriber to adjust the percentage of trade size that the system provider is trading. People don’t sign up for autotrade to have to continually manage the trade size for their account. (Please, let’s not turn this into an argument of how much attention people need to pay to their own account. I’m just saying, in general, people prefer to keep things simple.) If the account size is fixed, subscribers can at least count on something remaining constant from which they can base their setup.
-lcg
Thanks Matt. I didn’t realize it had been tried already. I understand some of the difficulties trying to add the feature you mentioned after the fact. My only suggestion would be to avoid making it an option for system providers. Just do it, run the calculation, and present both results, as it is now and with the simulated static available trading capital.
-lcg
Matthew: One problem with the current method is that if the portfolio grows to a large size, the system/vendor is forced into ever larger order sizes to maintain the same risk and profit relationships. As a result, order sizes can get to the point that fills are impossible and your Realism Factor punishes the system. If it were maintained at a smaller level, then order sizes would stay reasonable.
Perhaps there should be a way for the vendor to have a “Spit” in the value of their system, thereby normalizing it to a value that best fits them and their customers. Stocks do it. Why can’t we?
I’m assuming you mean “split” and not “spit.” When I first read your message I thought there was a really interesting and colorful financial term I had not heard of before.
