what's the deal with collective2?
- Thread starter IronFist
- Start date
In reality, it doesn't make much of a difference if the trades reported are sim or live. A subscriber who is autotrading will experience slippage regardless.
So, if the results the subscriber gets in his live autotraded account are reasonably close to the developer's he should be satisfied.
One could argue that slippage vis-a-vis the devoloper's reported fills are a small price to pay for not having to sweat the details of the system and just mimicking the signals, assuming of course the trades are profitable in the aggregate after costs, and better than the subscriber could get with other available investment opportunities, assuming equal risk.
I equate risk and reward relative to benchmarks. Like this year, NAZ draws down 60%, I do 33%, I still have a 47% return since 3/20/2007.
Let's say the case was that the market only drewdown 10%. That'd be 1.33^(1/3) about for drawdown and 1.47^(1/3) about for APR relative to the benchmark. All relative. If you're a market maker with a 10 million dollar portfolio, it could go down the next day by half, so it's not necessarily true that you won't be correlated with the market.
No doubt; however, the average Joe who is a subscriber to C2 can't avail himself to the strategies and commensurate returns of a market maker, so the S&P is arguably the best benchmark to measure performance against the C2 system developer population.
According to the sample distribution I posted earlier, the average return for the collective systems posted was -24%/week with SR ~-.4. Not exactly on par with S&P500.
However, that being said, the samples I took were from the extremes (that we have available). And I'm still not clear if those samples are from actively traded systems or continuously monitored archived systems from the supposed population of 7000 + systems (which seems to have been debunked here; in which case, are the top +/-15 weekly lists taken from actively traded systems?).
So far the track record length seems to be max 3 yrs? It would be nice to also have them show at least a simulation of back-tested results of the high performing systems for much longer lengths of time. Otherwise, with such a wide range of variance in system performance, we have no idea how well the top 5% winners might be expected to perform OOS.
