Serious Problem Continues in C2's System Accounting

I posted before about a serious problem in C2’s system results reporting. It was obvious to me that something was wrong because I was autotrading and had more equity than the model futures system I was trading–yet I was unable to take trades that the system took because I didn’t have enough margin to take the trades. Part of the problem was that C2’s margin table was wrong, but that didn’t explain the whole problem. And that problem continues.

There’s a perfect public example now in the trades of PatternZ. You can see a discussion of it in another topic, but I wanted to single out the reporting problem for discussion.

On 12/3/2015, the system went short 16 contracts of TF Dec 15, the e-mini Russell contract. Initial margin per contract is $6750; maintenance is $5400. The model account would have to have had $108,000 in equity to take the trade and $86,400 to hold it overnight. Yet at the close 12/2 the equity was $16,456 and had dropped to $5349 on the close 12/3. That was the only trade held from 12/3 until it was covered on 12/8. Your own stats say the equity on 12/4 was NEGATIVE $8805, 12/5 NEGATIVE $10651; 12/6 NEGATIVE $9577; $12/7 NEGATIVE $2602. Yet you report that the trade was closed on 12/8 for a profit of $33,217!

Clearly, C2 has a problem.

No broker would have allowed the system to go short on 12/3 in that quantity. No broker would have allowed the position to remain open as the equity went negative.

I’d love to trade with no money until the market recovered and I could claim a huge profit in my trade. My broker won’t let me do it, alas.

By the way. I know the model account includes fees and commissions, and the fees are not paid from the brokerage account. But The system has been trading for 6 months at a fee of $199 per month, so the fees don’t explain the problem


While we are on this topic, I will point out that C2 also doesn’t properly reflect realistic buying and selling prices for OTC equities.

While unrealistic from a trade volume perspective, filling the buy orders at the real time offer and filling the sell orders at the real time bid would at least be semi-realistic.

In the case of simulated trade executions for OTC equities [or anything else with a large bid/ask spread]: for limit orders, C2 should fill a buy limit order only when a trade occurs in the executable price range AND the offer price is at or lower than the buy order price, and C2 should fill a sell limit order only when a trade occurs in the executable price range AND the bid price is at or above the sell order price; for stop and market orders, I assume that C2 would fill buy orders only at the offer price and would fill sell orders only at the bid price.

For a prime example of this, look at this system which is currently ranked in the top 5% of systems on Collective2 (with a C2 score of 99.7):

I am curious to know what this system’s C2 score would be if the buy and sell prices actually reflected selling the large share quantities at the bid price and buying at the offer price. Would the system’s returns even be positive?!

The largest profit from Mehta Biotech Capital is a trade done on TROXG for a profit of $82,979. This trade is not even possible based on the price history on TROXG. It says it took a trade @ 0.28 and sold the next day for $20+. Definitely a problem. I saw something similar a few weeks back on a system that was trading crude oil futures and the system shorted CL @ $40 and covered hours later @ $20. This price action was not possible and should be investigated on how a system is able to fill trades that are not theoretically possible. System developers such as Mehta Biotech are exposing this flawed logic and making their systems appear to be the holy grail, and in actuality they will fail with real life orders.

These are serious problems and thank you both for highlighting them. Having dealt with the C2 team for several years I am sure they will take them seriously and correct the issues as soon as they can.

Lets just remember that financial data has always been and will always be a very problematic data source to stay on top of. And whilst C2 must be diligent to catch the errors it is a big help for them if we the community can continue to highlight problems as soon as we see them. So thanks again.

Any system developer who knowingly allows a beneficial error to go unnoticed is a cheat, plain and simple. An aggressive response would be to kick them off the network for good. Of course it is sometimes hard to say whether they knew or not whilst at other times it is as obvious as the nose on your face.

For now lets keep it civil and refrain from accusation. Instead lets just ask the developers directly:

@JohnPelsky of Pattern Z - Do you agree that these fills are fictitious? Have you written to to have them corrected?
@NiravMehta of Mehta Biotech Capital - Do you agree that these fills are fictitious? Have you written to to have them corrected?


Hi Matt, I have a suggestion that any system that goes into negative equity should immediately be frozen. You can see from the chart of Pattern Z that it was way underwater.

yes it is HORRIBLE and if its not fixed soon I am not sure how long i will be around especially as a developer. People see stats like this and they want to run to these programs due to the lure of quick money and big gains. It takes away from the good systems on here that use good money management and place realistic trades per the margin AND stated starting account size.

its simple in real life when you hit your margin requirement you get immediately pulled out of the trade and then in order to trade that same system the same way with C2 you should have to start over. the bigger problem is also that OLD records NEED to stay there for the same trader tracked by their tax information or similar. By law, Im pretty sure you have to post old track records even if they are not still in use to show people what you did before. You can say its a new system with a different method but you still have to post it so people know what you did in the past, even if new strategy is different. C2 needs to check with their compliance team on this. We deal with huge brokers and that is the way it has always been.

There are a few issues here. I’ll address them in separate posts.

First, let’s talk about systems that show equity curves dipping into “negative territory.”

The way C2’s equity curve works is that we first calculate the trading-results only (i.e. profits/losses from each trade). Let’s call this the “pristine” equity curve. Then, in a second phase, we “overlay” various costs that need to be applied. These costs can include things like: System Subscription Prices, AutoTrading Fees, Broker Commissions.

The reason for this two-step methodology is that it allows us to be flexible and show different commission rates to different people (perhaps you live in a country with one set of brokers rather than another). Or it allows developers to change their subscription fees, and have the equity curves adjust ex post. Or it allows users to select a “C2 Preferred” broker and see how the per-trade fees decrease, but the monthly trading fee increase. Etc.

In other words, it gives flexibility and allows users to visualize different scenarios.

The downside is that, under some of these scenerios, it allows equity curves to show as “negative” numbers, which of course is impossible in a world of one single trading account trading one single strategy in the exact same notional dollar amount.

When you see an equity curve dip into “negative” territory, what it really means is that: if you include system fees, plus autotrade fees, plus commissions – using the combination of these that you have chosen to view – you would be underwater.

(Technically speaking, by the way, if the “pristine” trading-only equity curve ever goes down to zero, the system can’t open any new trades. So that part happens to mirror real life.)

I know that many people who post on these forums have an opinion, typically strongly felt and loudly voiced, about how they would do things differently and much much better when they write their own version of Collective2 software.

But let me suggest that perhaps more is being made of this than is deserved. The equity curves you see here are labeled as hypothetical Model Accounts for a reason: because they don’t purport to show any single real-life trading account, at the quantities shown. These are “Model Account” performance records.

To help think about what this means: Imagine a trading system with a notional Model Account size of $100,000. When portrayed on C2’s"default view" (i.e. expensive commissions, full autotrade fees, full system subscription) it goes into negative territory.

A diabolical shortcoming? A horrifying mistake?

Well, maybe. But what if someone traded the strategy with a $200,000 account? He’d still be able to trade it, even though cumulative results of the system, including all fees, are negative. Or what if he trades two strategies in his broker account – one profitable and one losing? He still might be able to trade the system in question.

Or what if he used a different broker, with more favorable commission rates? He’s still be able to trade it and the curve would look different.

The point is:

This isn’t meant to be trading simulation game, one where we give people a virtual account balance and try to mirror the way a single real-life account works. Rather, the point of C2 is to portray the results of trading strategies in a flexible, accurate way - a way which allows people to make decisions about whether a strategy is worth considering.

So, if there is a customer out there who looks at a downward sloping equity curve that goes negative… and he still wants to trade the system, presumably that’s because he thinks there is something about that system worth trading.

So I encourage all of us to take a breath. Let’s not get too wrapped up in this issue that results can be show as “negative” here on C2. That doesn’t imply you can open a broker account and trade with a a negative balance. Rather, it means the system lost money, when all costs were applied to it. It’s still potentially tradable (in a larger account, or at a different broker, or at a lower subscription cost, etc).


Uh, NO. I considered that scenario–it doesn’t apply at all to the problems in my own autotrading and to the example I gave. Look at my original post. At the end of it, I said:

“I know the model account includes fees and commissions, and the fees are not paid from the brokerage account. But The system has been trading for 6 months at a fee of $199 per month, so the fees don’t explain the problem”

In the example I gave, on 12/5 the equity was negative by over $10K and the maintenance margin to hold the contracts was over $85K. Are you really saying fees were over $95K? And if you’re allowing systems to use more money than the model accounts state they have, why do you claim that the model account reflects (with minor caveats) roughly what the system follower will see? Your performance record is calculated on the initial deposit stated. If system developers can suddenly make hundreds of thousands of dollars appear to trade with after the system starts–with no mention and no indication of that whatever–shouldn’t you at least show the added funds and recalculate the theoretical system performance based on the total capital put in?

Here’s what the model system performance is based on according to C2’s own explanation:

"Material assumptions and methods used when calculating results
The following are material assumptions used when calculating any hypothetical monthly results that appear on our web site.
Profits are reinvested. We assume profits (when there are profits) are reinvested in the trading strategy.
Starting investment size. For any trading strategy on our site, hypothetical results are based on the assumption that you invested the starting amount shown on the strategy’s performance chart.

In some cases, nominal dollar amounts on the equity chart have
been re-scaled downward to make current go-forward trading sizes
more manageable. In these cases, it may not have been possible
to trade the strategy historically at the equity levels shown on the chart,
and a higher minimum capital was required in the past.All fees are included. When calculating cumulative returns, we try to estimate and include all the fees a typical trader
incurs when AutoTrading using AutoTrade technology. This includes the
subscription cost of the strategy, plus any per-trade AutoTrade fees,
plus estimated broker commissions if any."

Note the part I bolded…

And please notice that if this sort of trading is allowed, you’re clearly not considering the margins required for trading at all–which is what you said you certainly did consider when I brought the same issue up last June in another thread. What’s the point of the margin tables when they’re not being used at all to decide what trades can be taken in the account? Obviously they’re meant to be used in your calculations–and obviously they’re not being used in the proper way.


Jeffrey -


You are right about this one. It looks like C2 allowed the trade due to day-trading margin ($660 per contract) but did not impose the overnight margin of $6750 at the end of the day.

I need to look into why we didn’t issue a simulated margin call after market-close, which would have closed this position.

Thanks for the heads up.



Thanks, Matthew. I’ve always been impressed by your responsiveness on C2–it’s one reason I’m here.

When you delve into the problem, I think you’ll see it isn’t only for the example I gave or systems that go to negative equity–it applies as well to other systems which have a considerable amount of equity, but not enough to support the trades that are taken and kept overnight. I’ve had positions liquidated by my broker because I didn’t have enough equity to hold positions overnight, while the model system I was following had much less equity than I did but blithely held its positions overnight and later realized good profits on them as I observed the losses on the same trade in my account because I couldn’t keep holding it.

1 Like

Matthew, I must agree with Jeffrey. I don’t think the problem is only with overnight margin but with margin generally. I take Draken EFT example again. When I joined the system, I allocated the same account value as was the model account but to be on the safe side I scaled down to 70 %. Pheh, it was far to much. At the end I had to scale down to 25% to be able to follow.

He, with 50K account could buy 3x leveraged ETFs worth of 100 - 130K. How is this possible in a real life? IB gives you only 1,1 leverage on 3x ETFs (with margin account, not with portfolio margin account) so in the best he could buy 55 - 60K (I’m not 100% precise but for the illustration it is ok).

So can C2 explain how can a system owner trade with such leverage if broker doesn’t allow it? I’m talking specifically about IB, not sure about the others.

Matthew, will all the respect, this explanation is far from the real situation. A model account value -86K from 50K can’t be explained by fees. My only explanation is what I wrote above - the owner was allowed to use more money than in real life - either due to the wrong margin calculations or for some other reason. He still holds the positions irrespective the massive DD. And when the trend changes he could recover - but this is only a fairy tail.

At the end, showing that equity curve is unrealistic, wrong and misleading.

1 Like

Actually, we don’t take into account the fact that some levered ETFs require stricter margin requirements.

So that’s an issue I need to fix asap. Right now, all stocks and ETFs are treated the same in C2’s simulation of margin: you are allowed to trade with 2:1 leverage (i.e. put up 50% of the purchase price).

I’ll need to make a list of ETFs where only 1:1 margin is allowed.

If anyone can point me to an online resource showing special ETFs with unusual margin requirements, that would be helpful.

1 Like


Here are three links that will help in adjusting the margin requirements for double and triple leveraged Exchange Traded Products:

Hope this helps!



One more useful page from IB: (see at the bottom of the page for Leveraged ETFs).

Thanks for your input on triple-leveraged ETFs. I have added new special margin requirements for C2’s Model Account margins to better mirror the way IB handles it.