You are confusing ETF Capital strategy with Quantiger.
1 Like
Yeah I noticed this but wanted to stay out of it.
Here’s a thought, how about C2 implements a new category,
Systems that have endured bear markets and how many they have made it through. What do you think Matthew…neat idea!
2 Likes
I’d go a few steps further.
To be part of C2, each system has to be evaluated and vetted by C2 prior to being allowed in the C2 ecosystem?
There should be some sort of risk disclosure section - where backtest and/or walk-forward performance for a given algo across historically bearish periods are displayed.
For backtested systems, the algo author should be required to run a one month simulation with C2. After a month, demonstrate to C2 that their backtest of the same period generates similar signals to that the simulation period generated. This should help screen out fraudulent systems that repaint indicators, have invalid look-aheads, etc.
Assuming the simulation period and backtest results match, additional performance charts and trade logs would be produced (utilizing bearish market periods) with the algo creator and C2 in person. Those graphs would be provided within the algo’s description page.
1 Like
Keep in mind that some markets (like Forex and some commodities) have no “bear” cycle.
And you are obviously talking about “long only” systems, because a profitable trend-following system can make money in up and down markets.
After all, a bear market is just a downtrend that you can also ride/short.
1 Like
all the volatility that came from march 2020 created the most profitable trading environment myself and many others have ever seen in our lives (which really extended to end of march 2021).
not necessarily a bad thing.
3 Likes
Exactly.
Who cares if the market is up or down, since we can profit from both trends?
1 Like
The bear cycle is descriptive of the baseline, S&P buy-and-hold, not the algorithm. Feel free to suggest a better baseline.
The assessment is always relative to our options. So the S&P over bearish period vs the performance of the algorithm over the same period has validity.
1 Like
So what do you mean by “historically bearish periods.” ?
1 Like
E.g. 2008 crash. March 2020.
1 Like
So you are basically saying that systems do not perform well during crashes?
1 Like
No, not at all. I’m not concerned with all the systems. I want to evaluate specific systems that meet my criteria.
That to fully evaluate an individual system, I want to view performance numbers across a multitude of market dynamics, based on a typical baseline. The S&P is fine.
1 Like
C2 business model is to set connection between the trade leader and person with the money. This is it. If they would want to make an investment decisions and/or recommend the trade systems, than they would be a hedge fund. If investor can’t select the system here with all data C2 provides, then he should hire finance advisor.
There are systems that can’t be backtested. Should they be banned from C2 forever?
3 Likes
That info (price of the S&P) is clearly visible on the equity curve of each system.
1 Like
The S&P is visible for the time period that the system has traded under C2.
Additional performance data does exist for those systems that are back tested and automated. And that should be disclosed.
1 Like
Backtested performance data includes all of the numbers that are currently provided in C2 statistics, they’re just for alternative time periods.
Non backtested systems are fine. They should be disclosed as such. That distinction is just another data point in the decision making progress for any given investor.
1 Like
This would penalize otherwise good discretionary trading systems, because they cannot be backtested in most cases (due to their subjective trading rules).
1 Like
I believe that you always can ask trade leader for the backtest results as well as repeat your request in X months and compare the performance on C2 and backtest results during these X months. If trade leader hesitates to provide this info, than you make your conclusion.
2 Likes
The individual investor makes that judgement. To many it wouldn’t matter.
Whether a system is automated or not is just a label. Just another data point.
1 Like
Agree. I can ask for that data. And I have done that.
We agree on the fact that backtested performance numbers exist for individual algorithmically developed systems. They exist for time periods outside of the time span the system is traded through C2.
We disagree on the point of whether that data should be openly disclosed or not.
I feel it should be disclosed, up front, and allowed to be part of the decision making process for every potential investor. The transparency is necessary.
You still have no guarantee that the backtest is real.
