Let me preface my complaint by stating that I think C2 is an excellent service, and Matthew was very responsive to my concern and called me personally. It was a pleasure speaking with him, although probably not mutual 
Facts:
I signed up for a well-known system on May 11.
I activated auto-trading AFTER HOURS at about 16:30 EST on May 14.
I logged in to C2 to check my trade status on the morning of May 15.
To my surprise, I had been filled for 2 positions that were signaled on May 14.
My two fills occurred 17 hours after the system entry signals.
As Matthew explained, the auto-sync feature of C2 attempted to fill me for any pre-existing open trades. Since I activated auto-trading after hours, my broker filled these orders at the open on the following day. Hence, I got filled about 17 hours later. Here’s where my story gets very sad:
For these two trades, the system P/L is +$1000 per the target entry.
For these two trades, my P/L is -$1800 (ouch) per my delayed entry.
Why? Two reasons:
Those trades, which were long, just happened to gap down in AH.
Because auto-synchronization filled me 17 hours later.
Admittedly, I was ignorant that this would or could happen. Hard lesson learned. But, reasonably, this should not have happened. If auto-sync can turn two large winning trades into two large losses, then auto-sync is seriously flawed software which needs to be fixed. I don’t think Matthew was willing to concede this point, and I was too emotional to articulate it very well.
In absence of a fix, to completely avoid this nightmare, here is what I would have done differently, and what I suggest anyone else do who wishes to auto-trade:
1. Remember: Auto-Sync will try to put you into any open positions.
2. BEFORE you activate auto-trade, check the trading system for open positions.
3. If there are any open positions, DO NOT activate auto-trade.
3b. If your system constantly has open positions, set the scaling to 1% to protect yourself until auto-sync is no longer trying to put you into OLD open positions.
4. Only activate auto-trade during trading hours.
I expect there will be some discussion about these rules. But if you do not follow such guidelines, you do not know how a pre-existing open trade will end up with respect to your late entry price, which is likely very different from the system’s entry price. No one in their right mind would want to get in on a trade that is far away from the target entry price. Thanks to the limitations of auto-sync, activating auto-trading after hours while the system has open positions, renders that system mere gambling, which is presumably what we are all trying to avoid.
The obvious fix to auto-sync is this: Make auto-sync work only for NEW entry signals. Put a reasonable time limit on it. During our conversation, Matthew seemed to think this is technically unfeasible. He did not ‘bore’ me with the details, and I did not ask for them.
Matthew: Please share your thoughts.
John
Dear John,
I observed similar things (although on a demo account) - also shared here in a thread.
Already at this time I considered this is a bug, as it goes against the interest of the subscriber AND the provider.
It simply should be fixed by only allowing new signals to be accepted (this is eg. what Gen2 software like trader68 does.)
Klaus
Klaus-
Thanks for confirming this as an obvious bug. Good to know Trader68 only accepts new signals as it should. Ironically, I have avoided Gen2 because of reliability issues (mostly: power supply, IP provider, and other ‘acts of God’).
John
I think your scaling approach will also cause issues when you try to scale-up. The way auto-sync works now is good for slow trading systems, but not for fast trading systems.
The solution (for Matt).
"Auto-sync on current trade after setting up autotrade: yes/no"
Make a parameter (like above) an option that a system vendor can set.
Also add, “delay time before conversion to market order”, a vendor option as well.
It’s best to set-up autotrade with zero current trades an none expected for 60min. There is a 20-60min delay for set-up with some accounts. This does not work for some systems.
Jim
Dear Jim,
I do NOT think this is about a vendor parameter:
if it is a parameter, it has to be customer-set: S/HE is going to take the risk, not the vendor.
But to have automatic behavior would be ok, if
a) only new trades are entered (actually with trader68 etc. you can set how old signals can be to be accepted - tpyically is 15 minutes)
b) or only market is entered if the trade is better or equal than at the original time of the signal.
Cheers
Klaus
Yes I agree, but the vendor knows what is best for their system, so they should be able to set the default parameter for their system.
Jim
Update: The two trades have now been closed out by the vendor.
Vendor P/L:
Stock A: -$83
Stock B: +$769
Total: +$686
My P/L:
Stock A: -$1263
Stock B: -$1059
Total: -$2322
This is very difficult to accept.
I’ve just completed a more thorough audit of my account (my broker’s account statements are not the easiest to decipher). My initial stats grossly under-estimated the extent and damage of today’s late fills due to auto-sync.
It turns out that auto-sync put me in 7 open trades this morning before running into margin limits, or else it would have tried to put me into two more. All 7 were losers for me, naturally. Gee, I wonder why. Could it be because the system vendor never intended those entry levels? Could it be because auto-sync is a retarded POS? Yes, I think that’s it.
Drum roll…
Vendor P/L:
Stock A: ($427)
Stock B: ($44)
Stock C: ($232)
Stock D: ($352)
Stock E: $769
Stock F: ($83)
Stock G: ($611)
Total: ($980)
As you can see, the vendor had an off day. These were all longs placed a day or more earlier that got caught in today’s sell off. No big deal. It happens. (Why beat around the bush: The vendor is Gulf Stream 10 SE.) But the vendor had nowhere near the off day that I had, thanks to auto-sync:
My P/L:
Stock A: ($436)
Stock B: ($350)
Stock C: ($271)
Stock D: ($126)
Stock E: ($1040)
Stock F: ($1242)
Stock G: ($166)
Total: ($3632)
So my ($1800) nightmare becomes a ($3600) nightmare. Un-f—ing believable. (Excuse my language. Rough day.)
You can perhaps understand why I was initially fixated on the two ($1000+) hits to my account and failed to even notice the other five nightmare trades. Those two losses would have been avoidable were it not for auto-sync’s 17-hour late auto fills. The other fills were a day late, and one was two days late, because that’s how old those positions were, before auto-sync arbitrarily put me into these positions and did its damage.
I cannot believe this sort of problem has not occurred before.
Matthew: There is no excuse for this debacle. This is garbage programming and should not happen. I may not be legally entitled to compensation, but C2 is ethically dubious if some redress is not offered.
John:
As I explained when we spoke on the telephone, the AutoTrade software behaved exactly as it was designed to behave. AutoTrade performed as follows: your brokerage account matched the trading system from the point in time at which you turned on AutoTrading. In other words, on a marked-to-market basis, your account matched the C2 system you are AutoTrading almost perfectly.
I understand that you want the software to have behaved differently: you did not want AutoTrade to make your account mirror the C2 system performance from the moment in time that AutoTrading was turned on. Rather, you wanted AutoTrade to begin following the system only after the system went flat in a particular symbol.
That’s a perfectly reasonable product-design request, and certainly something we have considered adding, and (hopefully) will add. But I do take exception to your lurid and overwrought account of your harrowing “nightmare” – when in fact the software behaved exactly the way it was designed to behave, and when in fact that behavior was described in great detail on a screen that you were asked to read carefully and acknowledge before you turned on the AutoTrade software (and which you did acknowledge.)
I mean no offense. Your product-design suggestions are very welcomed, and the team here at C2 is anxious to make C2 AutoTrading exceed user expectations. But I do want to ask for a little reasonableness on your part: you are complaining about software functionality that is a conscious design choice, not a bug – and, further, you are complaining about being surprised about functionality which is amply described on the site.
In any case, hopefully in the near future we’ll be able to add more fine-grained control over which positions get synched when AutoTrading is first turned on. In other words, I hope in the near future to add the new features you are requesting.
I understand Matthew’s position. The basis idea of auto-sync is:
-whenever the system has position X, subscriber’s account has position X
- whenever the system doe not have position, subscriber’s account does not position X
So when the system wins 5%, your account wins 5%. When the system looses 5%, your account looses 5%. This is exactly what happened. You suggest that auto-sync turned a profit into a loss, but this profit was made before you started autotrading. The system lost money after you turned autotrading on, so your account lost money too. This is no flaw, it is what is supposed to happen.
Matthew-
I have tried to find the ‘fine print’ for auto-trading, pasted below. To my mind, it is somewhat general. Please correct me if I have the wrong text.
The problem I ran into would not have occurred to me based on my reading of the below. And it would not have occurred to me that auto-trade would enter me into ‘old’ trades, since that this is arbitrary and can be risky, as my example has shown. Anyway, we’ve been over all that. Perhaps this is a matter of my expectations not being met, as you (and now Jules) suggest. I concede your point that the software did exactly what it was designed to do, and that my dissatisfaction does not constitute a bug.
I do wish there were a better way of activating auto-trade so that the problem I have encountered could be avoided. As I have said to you and on this forum, had I known I’d be thrown into positions that were days old, I would have found a work-around, such as initially setting scaling to 1% until old trades were closed, because as a trader I consider such late entries to be unnecessarily risky. In this case, that risk was realized. If these trades had happened to go my way, I’d still have the same concern about auto-sync of old positions.
I do hope that future versions will provide the option to activate auto-trade for new trades only. That way, you sync into the system as new trades are put on. To my mind, synchronization should not just mean being in the position when the vendor is in the position no matter what; it should also mean getting into the position as close to the same time and price level as the vendor signal.
Finally, to C2 members: I hasten to add that Matthew has been responsive and very protective of his efforts and his business in addressing my concern. And I mean that in a good way. It’s clear to me that he is working hard to make C2 as good as it can be, and he takes very seriously criticism and suggestions. I apologize if my frustration suggested otherwise. I remain a loyal customer – as long as I become rich. Joke.
John
***
(5) TI will make a commercially reasonable effort to maintain synchronicity (“sync”) between the C2 trading system that the Trader subscribes to and the Trader’s brokerage account. However it is impossible to achieve perfect sync, for a variety of reasons:
a. A limit order may be briefly “touched” and filled in C2’s hypothetical trading record, but not filled in the Trader’s real-life account.
b. C2, TI, and the brokerage used by the Trader may receive different prices for futures, stocks, options, and foreign exchange because they use different dealers and quote feeds.
c. C2 is a simulated trading environment. For this reason, hypothetical fills may be posted incorrectly, or posted late, precluding perfect sync.
d. Software errors may occur.
e. Internet outages, power outages, computer malfunctions, and other infrastructure problems may occur.
(6) The list of potential causes of sync failure (above) is not exhaustive. There may be other reasons for potential failure. The Trader recognizes this.
(7) Because sync is not possible to maintain at all times, TI will need to make a commercially reasonable effort to restore sync between the Trader’s real-life brokerage account and the C2 trading system. These efforts will involve buying and selling financial instruments, typically at market prices. This may drastically affect the profitability of the trading system, possibly decreasing its profitability or increasing its loss.
(8) Typically there is a trade-off between maintaining sync and system profitability. Efforts to maintain sync typically reduce profitability.
The text I think you are looking for (which is displayed and must be acknowledged only the first time you configure AutoTrade) is this:
------
You must read and acknowledge the following before you can begin AutoTrading:
C2 uses “AutoSynchronization” technology to keep your brokerage account synchronized with the portfolio of C2 trading systems that you have chosen to AutoTrade.
The trading system called [SYSTEMNAME] currently has positions open. Because of C2’s Auto-Synch technology, Collective2 will place market orders (as soon as the market is open) to match the positions that are currently open in [SYSTEMNAME]. This means you will receive the current market prices to open these positions.
From a purely mathematical point of view, this does not change system performance. You are simply “joining late” a trade that is already in progress. You will achieve the same marginal results, from this point forward, as all other
AutoTraders receive from this point forward*. (*However, because these will be market orders, you may have higher slippage on these trades.)
Also, please note that, if your brokerage account currently has any open positions that are not part of a C2 system, C2 will close them in its initial attempt to get your account "in synch."
AutoSynchronization technology is designed to protect you to the extent feasible (although no protection is perfect).
We want to alert you about these potential trades before we place them, so that you are not surprised if you see them placed.
If you understand this, and are ready to proceed, click the button below.
Hi Matt,
well, I believe it is stated there clearly and precisely.
So, from this side it is clear.
However, you argued in an earlier post that this is not a bug and the software is working as designed. Being a professional in software quality, just a comment from my side: many (extreme) cases of software desasters are not cases of "bugs", but cases where the software is broken by design. I consider this as such a case, so the fact that it is "bug"-free does not improve the situation.
(I am actually considering to add this to my class material 
- just joking)
I was actually also just thinking about opening another account for full (FOREX) autotrading (so far I only used trader68), but now I have second thoughts about it.
It is not a problem of C2, but probably a fact that:
- most systems, especially in the FOREX area, do have at any time several overlapping trades (meaning, at no point in time there are no open trades),
- in cases like the "free" FOREX autotrading, even after subscribing one does not see what the current open trades are, without executing them.
- most, if not all systems accept severe, intermediate fluctuations. So you might get in at a much worse situation, if you enter intermediately.
- the approach for carefull entering is not really a solution. It only implies you do not take all these drawdowns at the same time, but distribute them over time.
- finally, this problem does not only occur when you first open an account, rather whenever you change the system allocation, e.g., if you change the systems traded in an account or only their relative percentage. Making this a constant (instead of initial) nightmare.
So, while I also made other proposals for features here on the website,
I would regard this is the most urgent feature request around, as it directly translates into a significant change of monetary performance on the subscriber side: and this is what C2 is all about.
Proposal would be to either:
- enter trades only for positions that are currently trading at a worse or equal level than as initial open (in this case the subscriber performance will at least not worse than the system performance).
- or not enter traders for which signal is already past.
Cheers
Klaus
P.S.: this is not about bashing C2, C2 is a major achievement and a lot of very good work has been put into it. But here something very simple can be identified which would give a GREAT bonus to the baseline of subscribers who want to do full autotrading.
The flip side is, of course, had you missed out on profits due to the software not synchronizing you with the current open trades, you’d be fuming about that “loss”.
Synchronization works as it should, in my opinion. Your account tracks the equity curve line of the system from the point you turn on auto trading. That’s the point of auto trading. Why would you want a different equity curve than the system has?
I’ve just completed a more thorough audit of my account (my broker’s account statements are not the easiest to decipher). My initial stats grossly under-estimated the extent and damage of today’s late fills due to auto-sync.
There is only one real problem here. Why were you using something new to you (which you did not really understand) with a live account and with so much money???
It is like a 12 year old child who decides that it is time to start driving. They have a key and a car, without their parents knowledge. Is Dad going to complain to the manufacturer about all the damage caused?
Hi Bundle Trader,
I think, one of the key points here is: no system gives signals that should be followed at an arbitrary later time.
From the perspective of signal provider you simple enter when the signal is given.
When you subscribe at an intermediate time to a system, you do not see the P/L of a system at this very minute. (This does not seem to be realtime computed at C2.)
Otherwise the advice would be: only enter a system, if the current P/L is largely negative. (In this case you do not get out worse than the system.)
The underlying problem is that many system only realize some small part of the the intermediate positive P/L. This is ok, if you entered early, but if you enter late, you are just sliding down the equity curve without the initial rising.
Cheers
Klaus
For what its worth, when I introduce I new system into live trading, I do what Klaus is saying:
1) If there are open positions, and they show a loss, I enter the positions when I start "live."
2) If there are open positions, and they show a profit, I do NOT enter the positions when I start "live."
I realize some may want to do exactly the opposite, and I understand why, but for me it is about reducing risk. If a position has profit already, my thinking is that the possible incremental profit to come does not justify the risk of a pullback.
I, for one, plan on telling my new subscribers to watch out for this (be aware of it, take whatever action you think is appropriate), although since the ES trades round the clock, my warning may not make it in time.
Kevin
Matt-
Thank you for posting the user agreement for auto-trading. I should have read it more carefully. I can see why you defend the software for “doing exactly what it was supposed to do.” It does. I should have exercized more caution when activating auto-trading.
You say:
"From a purely mathematical point of view, this does not change system performance. You are simply “joining late” a trade that is already in progress. You will achieve the same marginal results, from this point forward…“
Ah, but from a mathematical/financial POV, it can severely impact the subscriber’s portfolio, as we have seen. “‘Simply’ joining late a trade that is already in progress” gives different results for the subscriber than for the system, as I show above, for the same 7 positions, where the system lost $980, but I lost $3600, due to late joining via auto-sync. The potential damage depends on the nature of the system – the # of open trades, the size of the positions. The system to which I subscribed was invested ~$100k in the market at the time I got auto-synced in. If it had been a system that trades less aggressively, the marginal results, as you call them, would have been less risky, but still unacceptable to me, b/c arbitrarily entering a trade late is akin to gambling. That is why I advocate the scaling work-around I mention above, in lieu of a software option.
Klaus:
”- most, if not all systems accept severe, intermediate fluctuations. So you might get in at a much worse situation, if you enter intermediately.
Exactly. I hope my experience can help prevent others from suffering similar losses, if they fail to carefully read the instructions, as I did.
John -
It is indeed sad what happened in your case, and I applaud the fact that you are now trying to help/warn and bring it to the attention of others, so that in the future maybe this will not happen to someone else.
It won’t help you financially, but I bet somewhere down the line a subscriber avoids a loss from reading your story.
Kevin
Kevin-
I appreciate it.
I see your usual stellar rating has again inexplicably dropped from ~950 to ~550. Hopefully, it is a ‘marginal’ result
John
